An institutional body, most commonly found in colleges and universities, responsible for supporting and overseeing an institution’s standards, policies, and procedures related to academic integrity. Academic-integrity committees and related honor systems have longstanding roots in higher education; some institutional honor committees date to the late 19th century.

Depending on the institution, an academic integrity council or committee may educate students and faculty about:

• expectations for academic work
• review or recommend academic-integrity policies
• hear or adjudicate cases involving alleged cheating, plagiarism, unauthorized use of artificial intelligence (AI), or other forms of academic misconduct
• consider appeals or recommend sanctions; and advise institutional leaders about emerging academic-integrity issues.

Membership often includes faculty, administrators, staff, and students. The structure, authority, and responsibilities of these bodies vary substantially across institutions.

Generative AI is putting new pressure on these longstanding structures. Colleges report substantial increases in academic-integrity cases associated with AI, while faculty and committees face new challenges in determining what constitutes unauthorized AI use and how misuse can be established fairly.

Some institutions are consequently reconsidering traditional committee-based adjudication, shifting certain cases to instructors, designated academic-integrity officials, or centralized offices, while placing greater emphasis on prevention, clearer expectations for AI use, and redesigned assessments.

An emerging concept describing a higher education institution (university, college) that continuously adjusts its programs, policies, technologies, organizational structures, and services in response to changing learner needs, workforce demands, technological advances, demographic shifts, and societal expectations. Rather than relying primarily on periodic strategic planning or incremental change, the goal of an adaptive institution is to continually respond to change while maintaining academic quality, student success, and its educational mission. This often means multiple “systemic” changes, such as redesigning academic programs more rapidly, expanding inflexible learning pathways, strengthening employer partnerships, using data to inform decision-making, incorporating artificial intelligence and other emerging technologies, and creating new approaches to serving learners throughout their lives.

Although the concept continues to evolve, it reflects a growing recognition that colleges and universities must become more agile as education, work, and technology become increasingly interconnected.

Refers to a system or program that autonomously performs tasks for a user or another system. Agents are more sophisticated than AI assistants.

AI chatbot assistants use conversational AI techniques such as natural language processing (NLP) to understand user questions and automate responses.  These non-agency AI chatbots are ones without available tools, memory, and reasoning. The non-agency chatbots require continuous user input to respond, can produce responses to common prompts that align with user expectations, but perform poorly on questions unique to the user and their data. Since these chatbots do not hold memory, they cannot learn from their mistakes if their responses are unsatisfactory.

By contrast, AI agents can learn to adapt to user expectations over time. The results include more personalized experiences and comprehensive responses. Agents can complete complex tasks by creating subtasks without human intervention and consider different plans. These plans can be self-corrected and updated as needed. AI agents assess their tools and use their available resources to fill in information gaps. Since AI agents often do not have the full knowledge base needed to take on all subtasks within a complex goal, they use their available tools to include external data sets, web searches, APIs, and other agents. After the agent obtains missing information via these means, the agent can update its knowledge base and reassess its plan of action, self-correct, solve complex tasks in various enterprise contexts such as software design, IT automation, code-generation tools, and conversational assistants. The agent uses advanced natural language processing techniques of large language models (LLMs) to comprehend and respond to user inputs step-by-step and determine when to call on external tools.

Refers to a period in which enthusiasm, investments, or market valuations surrounding artificial intelligence (AI) exceed the technology’s demonstrated capabilities, adoption, or current economic returns. In an AI bubble, investors, organizations, and the public may overestimate AI’s impact, leading to unrealistic or inflated expectations, speculative investments, and rapid market growth that later may be followed by a correction.

Typically refers to generative AI models, such as ChatGPT. Machine learning, another form of AI, is not considered part of any potential bubble.

Refers to the idea that artificial intelligence (AI) systems and humans can work together in ways that improve performance, productivity, decision-making, creativity, or problem-solving beyond what either could accomplish alone. AI complementarity focuses on how AI can support, extend, or enhance human capabilities rather than replace human workers. In these models, humans and AI systems often perform different but connected tasks based on their respective strengths. For example, AI systems may rapidly analyze large amounts of data, identify patterns, automate repetitive tasks, or generate drafts, while humans provide judgment, context, ethics, relationship-building, creativity, oversight, and decision-making. Rather than assuming AI will eliminate most jobs, complementarity models examine how work roles may be redesigned so humans and AI systems contribute together.

The term is increasingly used in workforce and labor market research, economics, business strategy, higher education and workforce development, human resources and talent development, and public policy discussions about AI and the future of work and is becoming increasingly important in discussions about job redesign, reskilling and upskilling, skills-first hiring, human-AI collaboration, workforce productivity, and organizational change.

Examples of AI complementarity include:

• Healthcare: Clinicians use AI to identify possible medical conditions while clinician makes final treatment decisions.
• Education: Faculty members use AI tools to help generate lesson outlines while instructors provide expertise, teaching judgment, and student engagement.
• Customer service: Workers use AI-generated summaries or recommendations during conversations with clients.
• Researchers: Use AI tools to organize and analyze large collections of information while interpreting meaning and implications.

Refers to AI systems that autonomously simplify and organize tasks and processes for organizations.  Within the next two years, digital workers are projected to become an integral part of how many companies operate globally. Currently, digital workers are being used as an AI-powered sales development representative to handle the entire outbound sales development process for a company; an AI-powered caller to handle both inbound calls and consented outbound calls at a company; and an integration of these and other AI agents. The vision is that digital workers will:

• Scale existing teams at companies, using AI workers to augment human sales development representatives to allow them to handle a larger volume of prospects and communications.
• Replace human sales development representatives entirely, freeing employees to move into more advanced roles such as junior account executives.
• Reduce hiring needs, such as hiring fewer people when people retire because digital workers can handle much of the load.
• Empower account executives to use the digital worker directly to generate their own pipeline, reducing the need for a separate sales development team.

Projected benefits of digital workers:

• 24/7 operation, efficiency, and scalability (enhanced productivity).
• Personalization and tone matching (AI workers are designed to align with brand and individual communication styles, making their outreach feel authentic and personalized).
• Integration among AI Workers (enables multi-channel communication strategies).
• Freeing up human workforce (allows humans to focus on high-value activities).
• Continuous Improvement (AI workers are improved continuously based on user feedback and performance data).

Challenges of digital workers include:

• Ethical considerations (whether to disclose the AI nature of these digital workers to customers
• Potential saturation (potentially oversaturating communication channels).
• Integration with existing systems (ensuring seamless integration with a company’s existing tools can be complex and require customization).
• Data privacy/security (sensitive customer and prospect data requires sound security measures and compliance with data protection regulations).
• Training/adoption (learning curve for human teams to effectively manage and leverage AI tools).
• Performance monitoring/quality control (continuous monitoring to ensure AI workers are performing as expected, not making errors that could damage customer relationships).
• Scalability of personalization (when the volume of interactions increase, maintaining personalization and avoiding generic-sounding communications).

Refers to the growing set of organizations, benchmarks, standards, and research efforts used to test and assess AI systems. As AI tools become more widely used in workplaces, education systems, employer hiring processes, and public services, there is increasing interest in ways to evaluate how these technologies perform in real-world settings. AI evaluation efforts examine issues such as model accuracy, reliability, safety, bias, transparency, and how people actually interact with AI tools in practice.

Unlike traditional sectors where “product testing” is centralized or regulated through well-established institutions, the infrastructure for evaluating AI is still emerging. Evaluation activities are currently carried out by a variety of actors, including technology companies that test their own systems, academic researchers studying real-world performance and impacts, independent benchmarking organizations that develop standardized tests for comparing models, and government agencies developing frameworks and guidance for responsible AI deployment.

Several types of organizations participate in this emerging evaluation ecosystem. For example:

• Benchmarking organizations: Develop standardized tests for comparing machine learning models and tracking technological progress.
• Government standards bodies: Publish frameworks to help organizations evaluate and manage risks associated with AI systems.
• Academic research centers: Study how AI performs in real-world environments, including how users interact with these systems and how outcomes vary across contexts.
• Independent research and policy organizations: Examine broader social, economic, and governance implications of AI deployment.

As AI becomes more integrated into education advising systems, hiring platforms, workplace productivity tools, and learning technologies, the development of credible and transparent evaluation systems is increasingly viewed as essential. These efforts help organizations understand how AI functions in practice, identify potential risks, and support more responsible and effective deployment of AI in education, workforce development, and other parts of the learn-and-work ecosystem.

See Topic Brief: AI Evaluation Ecosystem | Learn & Work Ecosystem Library

Refers to the growing sense of exhaustion, skepticism, or disengagement that individuals and organizations experience from the rapid pace of artificial intelligence (AI) developments. AI fatigue may result from:

• Information overload
• Frequent technology changes
• Unrealistic organizational expectations
• Repeated vendor marketing
• Concerns about job displacement
• Uncertainty about which AI tools provide meaningful value
• Pressure to learn and adopt AI technologies without adequate training, support, or time to adjust.

Refers to the policies, processes, roles, standards, and oversight mechanisms used to guide how artificial intelligence (AI) is developed, selected, deployed, used, and monitored within an organization or system.

AI governance establishes who is responsible for AI-related decisions, what uses are permitted or restricted, and how organizations manage risks related to accuracy, privacy, security, bias, transparency, accountability, and other potential impacts.

Effective AI governance may include approved-use policies, human oversight requirements, risk assessments, data protections, documentation and audit trails, procedures for reviewing AI-generated outputs, and rules governing when AI systems may act autonomously and when human approval is required.

As AI systems become more capable—including agentic AI systems that can plan and carry out multi-step tasks—governance increasingly extends beyond approving particular AI tools to governing the processes, decisions, data, and actions in which AI participates.

AI governance seeks to balance innovation and responsible use by creating clear boundaries and accountability while enabling individuals and organizations to benefit from AI.

See Topic Brief: Regulation/Policy in Artificial Intelligence (AI) | Learn & Work Ecosystem Library

Several terms have emerged in the evolution of artificial intelligence (AI) that describe how multiple agents (machines, humans, or both) interact, share information, and produce coordinated or collective outcomes. These terms are related but they are not interchangeable. They reflect different aspects of how intelligence is distributed and organized across systems.

• AI Hive Mind
  • Descriptive, non-technical term for a system in which multiple AI agents are interconnected to share information, learn from one another, and coordinate actions in real time.
  • The agents function as a collective intelligence rather than as isolated systems.
  • Emerged as a term late 2010s–2020s, influenced in part by cultural references such as Star Trek and the rise of interconnected AI systems.
• Swarm Intelligence
  • Formal area of study that focuses in how decentralized, self-organized systems (often inspired by biological examples such as ants or bees) coordinate behavior through simple rules and local interactions.
  • Emerged as a term in the 1990s.
• Collective Intelligence
  • Intelligence that emerges from the collaboration and interaction of individuals or agents.
  • This concept applies broadly across human systems, machine systems, and hybrid human–AI environments.
  • Emerged as a term 2000s – 2010s as expansion of collective intelligence occurred across social, digital, and organizational contexts.
• Multi-Agent Systems (MAS)
  • A technical field in computer science focused on systems composed of multiple interacting agents, each with some level of autonomy.
  • These agents may cooperate, coordinate, or compete to achieve individual or shared goals.

These terms describe different dimensions of a similar phenomenon. Multi-Agent Systems refer to the technical structure of multiple interacting agents. Swarm Intelligence refers to a specific model of decentralized coordination. Collective Intelligence refers to the emergent outcome of group interaction. AI Hive Mind is a modern, metaphor-driven term that describes these systems as operating like a unified intelligence.

Use of these terms has implications especially for policy and governance, employers and workforce, and education:

• Policy and governance
  • Distributed decision-making complicates accountability and oversight
  • Transparency becomes more difficult as systems become more interconnected
  • Standards for interoperability and communication across systems become more important
• Employer and workforce
  • Movement from individual AI tools to coordinated networks of agents
  • Growth of roles focused on orchestration, supervision, and system design
  • Increased need for systems thinking and human–AI collaboration skills
• Education 
  • Shift from teaching students to use individual tools toward understanding systems of interacting technologies
  • Greater emphasis on systems thinking, including how multiple agents coordinate and influence outcomes
  • Expansion of digital and AI literacy to include evaluating outputs from multiple interacting systems, not just one tool
  • Emerging need to teach students how to orchestrate AI, including assigning roles to different tools, sequencing tasks, and monitoring results
  • New ethical questions around responsibility, transparency, and decision-making in distributed systems

Refers to the widespread promotion, enthusiasm, and elevated expectations surrounding artificial intelligence (AI) that may exceed its current capabilities or practical applications. AI hype is typically driven by media coverage, marketing, investment activity, public enthusiasm, and high-profile technology announcements that emphasize AI’s potential while sometimes minimizing its limitations or implementation challenges.

AI hype can accelerate awareness, experimentation, and investment, but it may also create unrealistic expectations about AI’s speed of adoption, accuracy, cost savings, or ability to replace human work.

AI hype is commonly associated with the early stages of emerging technologies.

AI literacy refers to the foundational knowledge and skills required to understand, use, and critically evaluate artificial intelligence (AI) systems. It typically includes basic awareness of how AI works, where it is used, its benefits and limitations, and its ethical and social implications. AI literacy emphasizes competent and informed use of AI tools. In most frameworks, AI literacy focuses on:

• Understanding what AI is and is not
• Knowing how AI systems are trained and deployed
• Recognizing bias, limitations, and ethical concerns
• Using AI responsibly and appropriately
• AI literacy largely assumes that users are interacting with AI as intended.

Adversarial literacy is an emerging concept that builds on AI literacy—but goes further. It emphasizes the ability to actively interrogate, challenge, and stress-test AI systems. It involves understanding how AI systems can be manipulated, misled, or produce harmful outcomes—and developing the skills to recognize, expose, and respond to those vulnerabilities. Rather than asking “How do I use AI well?”, adversarial literacy asks:

“How can AI fail, be exploited, or mislead—and how do I detect that?”

Adversarial literacy includes competencies such as:

• Recognizing adversarial inputs, prompts, or examples designed to manipulate AI behavior
• Probing AI systems to reveal bias, hallucinations, or unsafe outputs
• Evaluating AI responses for reliability, intent, and context sensitivity
• Engaging in ethical “red-teaming” or adversarial questioning to surface system weaknesses
• Understanding that AI systems are contestable, not authoritative

The term adversarial literacy is increasingly used by scholars and others to describe a paradigm shift, i.e. the emergence of several AI-driven shifts:

• AI systems are no longer passive tools: Generative and predictive systems actively shape information environments—and can be manipulated intentionally or unintentionally.
• Trust is no longer sufficient: AI outputs may sound authoritative while being incorrect, biased, or strategically misleading, requiring users to adopt a skeptical stance.
• Power and agency are at stake: Adversarial literacy equips individuals (earners, workers, citizens) to resist the over-reliance on opaque systems and reclaim human judgment.
• Education must move from use to critique: Teaching people how to prompt (chat with) AI is insufficient; they must also learn how to break, test, and question it safely and ethically.

See Glossary Term: Opacity (in Artificial Intelligence) | Opaque AI | Learn & Work Ecosystem Library

See Glossary Term: Information Literacy | Learn & Work Ecosystem Library

See Topic Brief: Converging Terms:  Digital Literacy & Information Literacy | Learn & Work Ecosystem Library

An artificial intelligence system (bot or AI avatar) deployed to attend a meeting or event in place of a human, typically to observe, record, summarize, or interact on their behalf. The “emissary” may be pre-programmed with instructions or agenda points, or be a real-time agent capable of basic interactions or just recording. The AI may appear as a name or video avatar in a Zoom/Teams screen and may or may not clearly indicate it is not human. The AI listens or records the meeting, may generate a summary or transcript afterward, and may or may not participate (e.g., say “noted,” “I’ll follow up.”  The human host of the meeting can typically recognize this is an AI representative but may or may not disclose this substitution to all other participants.

This is an emerging and controversial practice in AI-mediated communication that currently lacks a standardized term—but several terms are circulating in technology, business, and educational communities to describe the practice:

• AI Representative | AI Emissary
  • Not standard terms
  • AI Representative implies delegation (like a spokesperson)
  • AI Emissary has a formal, more diplomatic tone
  • Both suggest the AI is attending on behalf of a human, not pretending to be one
• Digital Double
  • Popular in speculative tech and AI circles
  • Suggests a full virtual representation of a person
  • Can imply real-time interaction or decision-making capability
  • Closer to Digital Twin or Avatar with Agency
• Ghost Attendee or Proxy Bot
  • Colloquial terms, sometimes used pejoratively
  • Emphasizes the ethical ambiguity—present but not present
  • Proxy bot may refer to a simple note-taking tool
• Shadow Presence
  • Describes situations where a bot is present in a digital meeting space (or even listens via device microphone), but its presence is not openly acknowledged
  • Often unintentional but raises ethical flags
• Synthetic Identity
  • An AI-generated name, avatar, or persona used to represent a bot in professional settings
  • Can be used transparently (e.g., “Fireflies.ai bot”) or misleadingly (e.g., a human-sounding name with no disclosure it’s AI)
• Meeting Clone or AI Meeting Attendee
  • Used in some productivity/startup platforms
  • Describes tools like Otter.ai, Rewind, Fathom, and Zoom AI Companion that show up with your name and join calls

Citation Note: This glossary entry was developed by the Learn & Work Ecosystem Library with support from ChatGPT (OpenAI, 2025) and is based on an original synthesis of emerging terminology from journalism, technical documentation, and academic discourse. The terms reflect evolving usage in the fields of AI, workplace automation, and digital communication, and are not yet standardized across sectors.

A tool that uses artificial intelligence to create resumes for individuals, from entry-level to executive level employment searches. The tool can write text, check the entire document for errors, and format the resume. The tool includes templates, writing tips, and automated features. Algorithms within the AI resume generator enables analysis of large amounts of data in order to provide individuals with tailormade content and design suggestions based on the user’s requests.

Since Applicant Tracking Systems (ATS) are increasingly used for first-stage resume review by many employers, there is growing pressure by job applicants to submit machine-readable professional resumes.

Examples of 17 of the best free resume builders recommended by C. Forsey at HubSpot’s  Marketing, Sales & Services blogs (April 2024):

• Zety: Best for Expert Resume Creation Tips
• Resume Genius: Best for Easy and Fast Resume Creation
• Wepik: Best for Customizing Pre-Made Resumes
• My Perfect Resume: Best for Guided Resume Creation Help
• Standard Resume: Best for Active LinkedIn Users
• Kickresume: Best for Quick and AI-Assisted Resume Creation
• Canva: Best for Design Creativity and Expression
• Pixpa: Best for Creating Online Resume Websites
• Indeed: Best for In-Platform Job Seekers
• com: Best for Minimalist Resume Creation
• Novoresume: Best for ATS-Friendly Resume Building
• VisualCV: Best for Multimedia Resumes
• CakeResume: Best for Resumes With an Online Portfolio
• Resume Now: Best for Time-Saving Resume Creation
• ResumeNerd: Best for Resume Writing Help
• Jofibo: Best for Comprehensive Guides
• Hloom: Best for Resume Templates

AI Study Companions and Intelligent Tutoring Systems (ITS) are both tools that use technology to help people learn, but they differ in how they are built and how they work with learners.

AI Study Companions are newer tools that use conversational artificial intelligence—like ChatGPT—to help students study. They answer questions, explain topics, and adjust to a learner’s needs based on what the learner says or uploads. These tools can feel like chatting with a helpful tutor and are easy to access on phones or laptops. Examples include ChatGPT Study Mode, Khanmigo, Duolingo Max.

Intelligent Tutoring Systems (ITS) are older, research-based systems built to teach specific subjects, often like a traditional tutor would. They follow a structured path, give feedback based on right or wrong answers, and track a learner’s progress over time. Examples include Carnegie Learning (for math), AutoTutor, ASSISTments.

A comparison of the tools:

Both tools aim to improve how people learn—especially by offering personalized help outside the classroom. AI study companions are becoming more common in everyday learning because they are easy to use and widely available. ITS tools remain important in formal education and research, especially where proven learning outcomes are needed.

Related Terms: Personalized Learning, Learning Technologies, Educational AI, Intelligent Learning Tools

A company or organization that develops, provides, licenses, or supports artificial intelligence (AI) technologies, platforms, software, infrastructure, or related services. AI vendors range from organizations that build large foundational AI models to companies that provide specialized applications, cloud computing services, hardware, consulting, implementation support, and system integration.

The AI marketplace includes several categories of vendors, each serving a different role within the AI ecosystem. Examples include:

• Foundation Model Developers – Organizations that develop large language models and other foundational AI systems that power many downstream applications (e.g., OpenAI, Anthropic, Google, Meta).
• Cloud and AI Infrastructure Providers – Companies that provide cloud computing environments, AI platforms, storage, and computing resources needed to develop and deploy AI solutions (e.g., Amazon Web Services, Microsoft Azure, Google Cloud).
• Hardware and Semiconductor Companies – Organizations that design processors, graphics processing units (GPUs), and specialized AI chips that enable AI systems to operate efficiently (e.g., NVIDIA, AMD, Intel).
• Enterprise Software Companies – Vendors that integrate AI capabilities into enterprise software used for business operations, education, workforce management, research, and productivity (e.g., Microsoft, Salesforce, Adobe, Oracle, Workday).
• Specialized AI Application Providers – Companies that develop AI solutions for specific industries or functional areas such as education, healthcare, human resources, cybersecurity, customer service, research, content creation, and scientific discovery.

AI vendors are becoming important partners in the learn-and-work ecosystem as rapid growth of AI has transformed the technology marketplace for educational institutions, employers, government agencies, workforce organizations, credential providers, nonprofit organizations, and other entities seeking to incorporate AI into teaching, learning, research, administration, human resources, workforce development, and other operational functions.

Unlike many earlier educational technologies that were purchased as stand-alone software applications, AI increasingly serves as an “enterprise” capability to support multiple functions across an organization. Selecting an AI vendor can involve broad strategic decisions about technology infrastructure, governance, privacy, security, interoperability, workforce readiness, and long-term institutional planning. Traditionally, many technology companies marketed products directly to individual colleges, universities, employers, and workforce organizations. While that model continues, AI adoption is increasingly occurring at multiple levels of the ecosystem. Depending on its intended use, AI solutions may be evaluated, negotiated, implemented, or supported through:

• Individual colleges and universities
• Multi-campus university and community college systems
• Statewide higher education agencies or purchasing agreements
• Regional (interstate) purchasing cooperatives and library consortia
• Enterprise technology providers and cloud platforms
• Professional associations and standards organizations
• Nonprofit intermediary organizations and collaborative initiatives

In some cases, a single agreement may support dozens—or even hundreds—of participating institutions. These collaborative approaches can reduce costs, improve security, simplify implementation, encourage interoperability, and promote the responsible adoption of AI technologies across multiple organizations.

As AI implementation becomes more complex, institutions are increasingly evaluating both individual AI products and how those technologies fit within larger institutional ecosystems that include existing enterprise software, data governance policies, cybersecurity requirements, accessibility standards, interoperability frameworks, procurement practices, and institutional AI strategies. AI implementation rarely depends on a single vendor. A college or university, for example, may rely on a foundation model developed by one company, cloud infrastructure from another, enterprise software from a third, and specialized educational applications from several additional vendors. Decisions about these technologies may occur at the institutional level, be coordinated across a university system, negotiated through statewide contracts or interstate consortia, or be influenced by enterprise technology partners, intermediary organizations, and standards organizations. As AI becomes more deeply integrated into organizational operations, institutions are increasingly managing interconnected ecosystems of vendors, technologies, governance structures, and partnerships rather than selecting individual software products in isolation.

Refers to a body-worn device (e.g., smart glasses, smart rings and watches, badges, or audio devices) that integrates artificial intelligence (AI) to collect data, interpret context, and provide real-time insights or assistance to the user. AI wearables may support functions such as learning, productivity tracking, health monitoring, navigation, language translation, and workplace performance support.

Within the learn-and-work ecosystem, AI wearables raise important considerations related to privacy, data governance, accessibility, human-AI interaction, and the evolving boundaries between personal technology and institutional systems.

Refers to the use of artificial intelligence (AI), machine learning, data integration, and automated analysis tools to support accreditation, quality assurance, compliance review, program evaluation, and outcomes assessment within education and workforce systems. These tools may assist accrediting organizations, institutions, workforce agencies, and policymakers in reviewing evidence, analyzing student and workforce outcomes, monitoring program performance, identifying compliance issues, aligning programs with labor market needs, and reducing administrative burden associated with accreditation and quality assurance processes—for both credit and noncredit programs.

AI-assisted accreditation remains an emerging concept, but it is likely to become a growing area of experimentation and policy discussion as education and workforce systems evolve toward more data-intensive, interoperable, and workforce-connected models. The concept is emerging alongside broader efforts to modernize accreditation and workforce education oversight in response to the rapid growth of:

• Nondegree and noncredit workforce programs
• Short-term credentials and Workforce Pell proposals
• Microcredentials and digital credentials.
• Skills-based hiring practices
• Apprenticeships and workforce-aligned education pathways
• Learning and Employment Records (LERs)
• Longitudinal student and workforce outcome data systems.

Several factors are fueling interest in AI-assisted accreditation and workforce quality assurance systems.

• Scale: Traditional accreditation systems were largely designed around degree-based higher education models and periodic review cycles. The rapid expansion of short-term workforce education, online learning, employer partnerships, and noncredit credentials is creating pressure for more scalable approaches to program evaluation and oversight.
• Cost: Accreditation and compliance processes are often labor intensive, document heavy, time consuming, and expensive for both institutions and accrediting organizations. AI-assisted systems are increasingly being explored as a way to reduce duplication, streamline evidence review, automate portions of reporting, and lower the administrative burden associated with accreditation and quality assurance activities.
• Demand for labor market alignment and outcomes transparency: Policymakers, employers, states, students, and workforce agencies are increasingly seeking evidence that education and training programs lead to employment, wage gains, credential portability, and career mobility. AI-assisted systems may help integrate labor market data, workforce information, employer demand signals, and student outcomes into more continuous evaluation processes.
• Growth of interoperable digital infrastructure: Increasing use of workforce data systems, digital credentials, AI analytics tools, and shared data environments is making it more feasible to connect education, workforce, and employment information across systems and organizations.

In workforce education specifically, AI-assisted accreditation is increasingly being discussed in relation to Workforce Pell implementation, workforce ecosystem coordination, statewide workforce planning, shared evaluator and reviewer capacity, regional labor market alignment, credential interoperability and portability, continuous quality monitoring, and AI-enabled policy implementation systems.

AI-assisted accreditation does not replace human peer review or institutional judgment. Rather, it generally aims to augment and streamline accreditation and quality assurance activities by helping reviewers process large volumes of information more efficiently, identify patterns or risks, automate portions of reporting and documentation, and support more continuous forms of monitoring and evaluation.

The growth of AI-assisted accreditation does raise important questions about data privacy, algorithmic transparency, bias, governance, institutional autonomy, reviewer accountability, and the appropriate balance between automated systems and human judgment

Refers to the laws, executive directives, guidelines, and strategic frameworks enacted by individual state governments in the United States to govern the development, procurement, deployment, and oversight of AI technologies. Not all states have implemented regulations /policy, but those that do typically address transparency, data privacy, algorithmic accountability, cybersecurity, civil rights protections, ethical use, and the responsible application of AI in public services such as education, workforce development, healthcare, and transportation. State policies may include requirements for risk assessments, impact disclosures, governance structures, procurement standards, and restrictions on high-risk or harmful uses of AI.

State AI regulations operate within a broader national framework shaped by federal AI policies—such as those issued by the White House, the United States Congress, and federal agencies including the National Institute of Standards and Technology (NIST). Federal policy establishes nationwide principles related to AI safety, national security, civil rights, and sector-specific oversight, while states develop complementary or more targeted rules that address local needs and contexts.

See: Learn & Work Ecosystem Library: Glossary / Artificial Intelligence (AI) Regulations / Policy – U.S. Government | Learn & Work Ecosystem Library

See: Learn & Work Ecosystem Library: Topic Brief: Regulation/Policy in Artificial Intelligence (AI) | Learn & Work Ecosystem Library

Refers to the laws, executive actions, regulatory guidance, and national strategies established by the United States government to ensure the safe, ethical, and responsible development and use of AI across the country. Federal AI policy sets nationwide principles and requirements related to civil rights protections, data privacy, national security, AI safety and risk management, transparency, consumer protection, and oversight in key sectors such as healthcare, finance, education, and transportation. Components of federal policy may include agency-specific standards, procurement rules for trustworthy AI, reporting and impact assessment requirements, and governance structures such as federal AI councils or interagency oversight bodies.

Federal policy establishes overarching guardrails and national coordination within which states may adopt more specific or context-based AI regulations. The U.S. approach is developing alongside AI governance efforts in other nations and regions—including the European Union, the United Kingdom, Canada, and member countries of the Organisation for Economic Co-operation and Development (OECD)—which are creating regulatory frameworks that emphasize safety, accountability, human rights protections, and international alignment. Together, these national and international approaches reflect a growing global effort to promote responsible and harmonized AI governance.

See Learn & Work Ecosystem Library: Glossary /  Artificial Intelligence (AI) Regulations / Policy – States | Learn & Work Ecosystem Library

See Learn & Work Ecosystem Library: Topic Brief / Regulation/Policy in Artificial Intelligence (AI) | Learn & Work Ecosystem Library

Artificial intelligence (AI) refers to the ability of a computer or computer-controlled robot to imitate human brain functions.  Machines use the application of computer science through algorithms to process large data sets to perform tasks typically associated with intelligent beings, such as the ability to reason, discover meaning, generalize, synthesize, and learn from past experiences.  Through rapid advances in computer processing, speed, and memory capacity, AI is used for more and more sophisticated applications such as medical diagnosis; computer search engines; voice, face, and handwriting recognition; and chatbots.

Generative AI refers to AI able to generate text, images, or other media in response to prompts. Generative AI models process large data sets of natural language, code language, and images to create new content in these forms (natural language, code language, images) and other data forms. Examples include ChatGPT, Bing Chat, and Bard. Many applications are using generative AI in the fields of art, marketing, writing, software development, product design, healthcare, finance, gaming, fashion, and education (in teaching, learning, student support services, and administrative supports).

AI prompts are any form of text, question, information, or coding that communicate to AI what response(s) are being sought.

Other terms for AI include machine learning (ML) and deep learning.

Refers to an institution of higher learning with an undergraduate population that is at least 10% Asian American and Native American Pacific Islander. This designation is defined under the Higher Education Act (HEA) and is granted by the U.S. Department of Education to support eligible colleges and universities in improving and expanding their capacity to serve Asian American and Native American Pacific Islander learners. There are approximately 200 AANAPISIs in 27 states and territories, mostly clustered on the west coast, in Hawaii and the Pacific territories, as well as New York.

Seeks to bring coherence to higher education systems across Europe. It established the European Higher Education Area (EHEA) to facilitate student and staff mobility, make higher education more inclusive and accessible, and make higher education in Europe more attractive and competitive worldwide. As part of the EHEA, participating countries agree to: introduce a 3-cycle higher education system consisting of bachelor’s, master’s, and doctoral studies; ensure mutual recognition of qualifications and learning periods abroad completed at other universities; and implement a system of quality assurance, to strengthen the quality and relevance of learning and teaching. Launched with the Bologna Declaration of 1999, the Bologna process is implemented in 49 countries, which define the EHEA. To become a member of the EHEA, countries have to be party to the European Cultural Convention and declare their willingness to pursue and implement the objectives of the Bologna Process in their own systems of higher education.

A technology system that enables direct communication between the human brain and a computer or digital device. Brain-Computer Interfaces (BCIs) use signals generated by the brain to control software, machines, or other technologies without requiring traditional physical input such as typing, speaking, or touch.

Some BCIs are invasive, requiring surgically implanted devices placed in or on the brain. Others are non-invasive and use external headsets, sensors, or wearable devices to detect and interpret neural activity.

BCIs are currently used primarily in clinical, medical, and research settings, including efforts to help individuals with paralysis, neurological disorders, speech impairments, or mobility limitations communicate or interact with digital systems.

The technology is also increasingly discussed within future-of-work, AI, defense, education, and workforce development conversations. Some researchers, companies, and futurists believe BCIs could eventually influence how humans learn, communicate, access information, interact with AI systems, or perform certain forms of knowledge work.

Although broad commercial adoption remains uncertain and likely years away, BCIs are increasingly viewed as part of longer-term discussions about human-machine interaction, cognitive augmentation, accessibility, and the future relationship between people and intelligent technologies.

An open standard developed by 1EdTech (formerly IMS Global Learning Consortium) that provides a structured approach to describing, collecting, and exchanging learning activity data at scale. This open standard defines a common vocabulary and data format for learning events, facilitating interoperability among educational tools and platforms. Caliper also specifies an application programming interface (API), known as the Sensor API™, for transmitting event data from instrumented applications to target endpoints for storage, analysis, and use.

Unizin integrates Caliper Analytics into its Unizin Data Platform (UDP) to collect and process learning activity data from various educational tools and platforms. By leveraging the Caliper standard, Unizin ensures that learning activity data from diverse sources can be integrated, analyzed, and utilized to enhance educational outcomes across its member institutions.

See:  https://learnworkecosystemlibrary.com/initiatives/unizin-data-platform-the-caliper-analytics-standard/

See: Unizin | Learn & Work Ecosystem Library

Refers to the practice of a higher education institution merging with another higher education institution facing closure typically due to severe financial problems. The institution being “bought” is often a private accredited institution. The new arrangements enable students from the closing institution to transfer their credits to the new institution and continue their pathway toward credential completion. An example is Northeastern University (Boston) entering into merger arrangements with 14 private, accredited schools as part of a trend of “chain buying” in the private school ecosystem.

See: Mergers & Acquisitions / Consolidations in Higher Education

The practice of using external tools, technologies, or actions to reduce the amount of mental effort needed to complete a task. Examples include using a calendar instead of remembering an appointment, a calculator instead of doing arithmetic mentally, or a search engine instead of recalling information from memory. The concept predates generative AI.

AI greatly expands the kinds of cognitive work that can be offloaded, including summarizing information, generating ideas, analyzing material, drafting text, solving problems, and organizing knowledge. Cognitive offloading can improve efficiency, but increased reliance on external tools may also affect how often people practice or develop the underlying cognitive skills themselves.

The intentional design and management of the linguistic, informational, and interactional context in which an AI system operates in order to influence its outputs over time. Unlike prompt engineering, which focuses on individual prompts, context engineering shapes the broader conditions that guide model behavior—such as persistent instructions, accumulated examples, role continuity, reference materials, and conversational history—leveraging how language models extend patterns across context windows.

Context engineering reflects a shift from controlling AI through phrasing to guiding it through structured environments, making it especially relevant for learning systems, knowledge platforms, and workplace AI tools.

See: Prompt Engineering | Learn & Work Ecosystem Library

Refer to AI-generated (also known as synthetic or manipulated media) which is often video or audio, that convincingly mimic real people’s likenesses or voices, making them appear as if they are saying or doing things they never did. These digitally altered materials are often generated through advanced machine learning techniques, especially generative adversarial networks (GANs), which can create realistic yet fake and misleading content.

In higher education, deepfakes pose significant risks, including threats to privacy, reputation, and information security. The rise of deepfake technology has led to serious cybersecurity threats, as seen in an incident at the University of Utah. Here, phishing emails were used to acquire legitimate photos of students, which were later manipulated into threatening deepfake images for extortion purposes. Such attacks represent a new form of cybercrime, leveraging AI-driven digital manipulation to compromise personal identities and reputations and, in turn, exploit victims for financial gain.

To combat these sophisticated cyber threats, many higher education institutions are adopting a multi-layered approach to cybersecurity that combines education, technology, and strict security protocols, to help campuses manage and mitigate these threats. Examples of strategies:

• Awareness and Training: Educating students and faculty about recognizing phishing scams and verifying the authenticity of emails, particularly those requesting sensitive information. Many campuses are implementing mandatory cybersecurity training programs to foster vigilance.
• Digital Literacy Initiatives: Incorporating digital literacy into curricula to educate students on the risks of AI, data privacy, and safe online behavior, especially as AI tools become more accessible and students themselves may experiment with these technologies.
• Email and Network Security: Upgrading email filtering and monitoring systems to detect and block phishing attempts more effectively, especially those targeting students with requests for sensitive information.
• Enhanced Authentication and Detection Systems: Investing in and upgrading multifactor authentication (MFA) and using anomaly-detection algorithms that spot irregularities in digital communications; e.g., inconsistencies in face movements, unnatural speech patterns, or other technical markers. These tools use algorithms that can identify alterations in the pixelation or lighting of images, helping to screen out manipulated content. Machine learning models can also help identify signs of manipulation in images or videos. Some universities are using digital watermarking and blockchain-based authentication to verify legitimate content.
• Incident Response Plans and Forensic Analysis: Developing specific incident response strategies for deepfake and cyber-extortion cases, enabling quicker containment, investigation, and, where necessary, cooperation with law enforcement. When deepfake-based threats or attacks occur, universities employ forensic analysis to trace the source of the attack and collect evidence. These efforts aid in understanding the vulnerabilities exploited and strengthening defenses.
• Collaboration with Law Enforcement: Collaborating with local and federal law enforcement agencies, especially in cases involving extortion or blackmail. This collaboration aids in swift actions needed to protect students and staff and may serve as a deterrent to future attacks.

The ability to use digital technologies confidently, adaptively, and creatively across changing contexts. It goes beyond basic digital literacy or task-specific competency by emphasizing judgment, problem-solving, innovation, collaboration, and the capacity to learn new tools as technologies evolve.

A digitally fluent person not only uses technology, but understands when to use it, how to combine tools effectively, how to evaluate outputs, and how to adjust as systems change. Digital fluency is increasingly important in workplaces shaped by AI, automation, data systems, and continuous technological change.

Digital fluency is often seen as an advanced stage of digital readiness built on foundations of literacy and competency.

The term can apply differently, depending on context.  For example:

• For younger students
  • Focus may include device basics, safe online behavior, media awareness, collaboration tools, and foundational creation skills
• For college learner
  • Focus may include research tools, digital identity, online teamwork, data literacy, AI-assisted learning, platform navigation
• For working adults / professionals
  • Focus may include workflow systems, cybersecurity awareness, data interpretation, remote collaboration, AI tools, continuous upskilling.
• For older adults
  • Focus may include access to services, telehealth, communication tools, fraud awareness, digital confidence, and independence and inclusion.

See Glossary Term: Digital Literacy | Learn & Work Ecosystem Library

Refers to a person’s ability to take an active role in seeking, evaluating, producing, using, and questioning knowledge versus simply accepting information provided by others. In education, it includes learners taking responsibility for determining what they know, what they need to know, and how their learning can be evaluated.

The concept has become increasingly relevant in the use of AI because AI systems can quickly provide answers, explanations, and synthesized information. Maintaining epistemic agency means that the human remains actively involved in judging the information rather than treating AI-generated information as automatically authoritative.

Refers to news, analysis, or informational content that remains relevant and useful over a long period, rather than being tied to a specific, time-sensitive event. This content continues to provide value to readers or users months or even years after publication. Evergreen journalism is especially valuable in education, workforce development, and lifelong learning, where users seek guidance, explanations, or historical perspectives that retain relevance over time.

In an AI-enabled information environment, evergreen content requires structured storage, metadata tagging, and retrieval systems to ensure discoverability and accessibility. AI tools can also help maintain its relevance by suggesting updates or linking it to new developments.

Examples

• Explainers on emerging workplace technologies
• Historical analyses of industries or labor trends
• Educational “how-to” or skill-building articles

Fair use is a legal doctrine in U.S. copyright law that permits limited use of copyrighted material without permission from the rights holder for purposes such as education, research, scholarship, commentary, criticism, and news reporting. Fair use determinations are context-specific and rely on an analysis of four factors: (1) the purpose and character of the use, (2) the nature of the copyrighted work, (3) the amount used, and (4) the effect on the market value of the original work.

The growth of artificial intelligence (AI) has complicated traditional understandings of fair use, raising new legal and institutional questions about the use of copyrighted materials.  In digital learning, research, and AI-enabled environments, fair use plays a critical role in enabling access, innovation, and knowledge sharing while balancing the rights of content creators. AI complicates fair use practices because it raises unresolved questions such as:

• Whether training on copyrighted works constitutes fair use?
• Whether AI outputs are “transformative”?
• Who bears responsibility when outputs resemble copyrighted material?
• How should attribution, licensing, and compensation work at scale?

Courts have historically emphasized human purpose and transformation. AI introduces non-human intermediaries, probabilistic reuse, and scale far beyond traditional educational copying. This matters for educators, researchers, libraries, and publishers because fair use is no longer just about what humans copy: it is about what systems ingest, how outputs are generated, and how institutions manage risk, disclosure, and compliance.

A growing threat facing companies globally are jobseekers who are not who they say they are, using AI tools to fabricate photo IDs, generate employment histories, and provide answers during interviews.  An imposter candidate who is hired can install malware to demand a ransom from a company, or steal its customer data, trade secrets, or funds. Estimates are that by 2028 globally, the rise of AI-generated profiles will result in 1 in 4 fake job candidates. Some companies report seeing individuals using fake identities, fake faces, and fake voices to secure employment, even going so far as doing a face swap with another individual who shows up for the job.

To address this problem, there is an emerging industry of identity-verification companies to weed out fake candidates.

Refers to the current era of technological transformation characterized by the convergence of digital, physical, and biological technologies. Building on earlier industrial revolutions driven by mechanization, electricity, and computing, the Fourth Industrial Revolution is distinguished by advances in artificial intelligence (AI), robotics, the Internet of Things (IoT), cloud computing, advanced manufacturing, biotechnology, quantum computing, digital twins, autonomous systems, and data analytics.

Unlike previous industrial revolutions that primarily transformed manufacturing, the Fourth Industrial Revolution is reshaping nearly every sector of society—including education, healthcare, government, transportation, finance, and the workforce. It represents a transformation in how knowledge is created, work is organized, and value is generated. Many occupations are being redesigned, requiring workers to continually acquire new skills and adapt to evolving technologies. Educational institutions, employers, governments, and workforce organizations are responding by developing more flexible learning pathways, new credentialing models, and stronger alignment between education and employment.

The term was popularized by Klaus Schwab in 2016 through his book The Fourth Industrial Revolution and has since become a widely used framework for understanding the societal implications of rapid technological change.

See Glossary Term: Industry 4.0 | Learn & Work Ecosystem Library

See Topic Brief: The Fourth Industrial Revolution: Transforming Learning & Work | Learn & Work Ecosystem Library

Refers to AI (artificial intelligence) able to generate text, images, or other media in response to prompts. Generative AI models process large data sets of natural language, code language, and images to create new content in these forms (natural language, code language, images) and other data forms. Examples include ChatGPT, Bing Chat, and Bard. Many applications use generative AI in the fields of art, marketing, writing, software development, product design, healthcare, finance, gaming, fashion, and education. In education, uses are increasing in teaching, learning, student support services, and administrative supports.

Related terms include machine learning (ML) and deep learning.

Refers to strategies used to improve how content is discovered, interpreted, summarized, and cited by generative artificial intelligence (AI) systems which include large language models and AI-powered search and assistant tools. GEO builds on rather than replaces traditional Search Engine Optimization (SEO). While GEO is gaining importance, SEO remains foundational but is no longer sufficient—it is being absorbed into the broader information stack  with two key implications: (1) users increasingly receive answers instead of links; and (2) content visibility now depends on whether it is considered authoritative enough to be included in an AI response and whether it can be summarized accurately without distortion. SEO optimizes for indexing and ranking in search engines while GEO optimizes for selection, synthesis, and citation by generative systems.

See: Search Engine Optimization (SEO) / Key Word Optimization | Learn & Work Ecosystem Library

A globally interoperable micro-credential is a digital credential designed to be issued, stored, shared, and verified across different technical platforms and regional (including cross-national) credentialing systems.  Interoperability is achieved by aligning credential data models and cryptographic verification methods (codes that keep information secret) with multiple recognized standards so that credentials are not locked into a single proprietary ecosystem.

These micro-credentials address a persistent challenge in the ecosystem: credentials that cannot travel across institutions, digital platforms, or national systems. Emerging implementations such as the Velocert Globally Interoperable Micro-credentials illustrate how technical and regulatory standards can be combined to improve portability, trust, and long-term verifiability of learning achievements. The Velocert micro-credential supports simultaneous issuance in Open Badges v3 and European Digital Credentials for Learning (EDC) formats. This approach enables a single credential to remain verifiable in both global badge-based systems and in European digital credential and wallet infrastructures.

Characteristics of these micro-credentials:

• Standards-aligned — Support multiple credential standards to enable cross-platform portability
• Machine-readable and verifiable — Use cryptographic methopds to ensure authenticity and detect tampering
• Learner-controlled — Designed for storage in digital credential wallets where individuals manage sharing
• Cross-border recognition — Align with European and international credential frameworks
• Metadata-rich — Can include learning outcomes, workload, assessment criteria, and skill taxonomies

See Initiative: Velocert Globally Interoperable Micro-credential Initiative – United Kingdom | Learn & Work Ecosystem Library

Refers to the structured and dynamic interaction between humans and machines, particularly AI-enabled systems, in which each contributes distinct capabilities to perform tasks, solve problems, or make decisions. Rather than replacing human labor, machines in collaborative arrangements augment human judgment, efficiency, and insight; and humans provide context, oversight, ethical reasoning, and accountability.

Although the phrase has existed in in the literature for decades (as HCI / Human–Computer Interaction and in robotics literature), it is experiencing an updated understandings due to generative AI and autonomous systems:

• AI systems now reason, generate, recommend, and decide; they do not just execute tasks.
• Collaboration is no longer sequential (“human → machine” or “machine → human”) but it is continuous and interdependent.

As AI systems become more autonomous and generative, human–machine collaboration increasingly functions as a continuous partnership rather than a handoff between human and machine roles.

A related term is Machine-Human Workforce which focuses on composition and structure; Human–Machine Collaboration focuses on interaction and process.

See Glossary Term: Machine-Human Workforce | Learn & Work Ecosystem Library

See Topic Brief: https://learnworkecosystemlibrary.com/topics/machine-human-workforce/

An emerging term used to describe situations in which organizations present artificial intelligence systems, automated process, or technology-enabled service as more human, human-controlled, or human-like than it actually is. The term is used to describe efforts that may mislead users about the role, capabilities, oversight, or involvement of humans in a system, particularly in contexts involving AI and automation. Current usage generally falls into three categories:

• Making AI appear more human than it really is
  • Refers to presenting AI systems as possessing human-like capabilities, emotions, judgment, or understanding that they do not actually possess, thereby misleading users about what the technology can do.
• Hiding AI behind a human façade
  • Where organizations imply that a service is provided by humans when AI is actually doing much of the work.
  • Examples include AI chatbots that are not clearly disclosed as AI or synthetic avatars presented as real people.
• Claiming “human oversight” that is largely symbolic
  • Organizations emphasizing that humans remain “in the loop” to reassure stakeholders, even when human review is limited, ineffective, or largely performative.
  • In this sense, humanwashing uses the language of human judgment and accountability to make systems appear safer or more responsible than they are.

Refers to the transformation of manufacturing and industrial production through the integration of advanced digital technologies, automation, and connected systems. Often described as the industrial dimension of the Fourth Industrial Revolution, Industry 4.0 emphasizes the use of intelligent technologies to create more efficient, flexible, data-driven, and autonomous production environments. Industry 4.0 combines technologies such as artificial intelligence (AI) , Industrial Internet of Things (IIoT), robotics and autonomous systems, cloud computing, big data and advanced analytics, digital twins, cyber-physical systems, additive manufacturing (3D printing), advanced sensors, and edge computing. Rather than relying on isolated machines or manual processes, Industry 4.0 enables equipment, products, and production systems to communicate, exchange data, and make real-time decisions. Manufacturers can monitor operations remotely, predict equipment failures before they occur, customize products more efficiently, and optimize production through intelligent automation. Beyond manufacturing, Industry 4.0 principles are increasingly influencing logistics, supply chains, healthcare, agriculture, construction, transportation, energy, and other sectors.

For education and workforce development, Industry 4.0 has increased demand for workers who possess both technical and human-centered skills, including digital literacy, data analysis, systems thinking, cybersecurity, AI literacy, collaboration, and continuous learning. It has also accelerated interest in industry certifications, microcredentials, apprenticeships, competency-based education, and employer-education partnerships designed to prepare learners for technology-enabled workplaces.

Industry 4.0 is widely viewed as one of the primary technological drivers of the broader Fourth Industrial Revolution.

See Glossary Term: Fourth Industrial Revolution (4IR) | Learn & Work Ecosystem Library

See Topic Brief: The Fourth Industrial Revolution: Transforming Learning & Work | Learn & Work Ecosystem Library

More than 170 countries and regions have published national digital strategies, and more than 50 countries have also developed AI strategies.  Strategies typically include a focus on building and maintaining an architecture to help align government and public service approaches for data sharing, adoption of technology, and exploiting AI (artificial intelligence).  The layers of this architecture consist of intelligent sensing, intelligent connectivity, intelligent foundation, intelligent platform, AI foundation model, AI large model, and intelligent application.  These layers are designed to help provide more inclusive and people-centric public services to promote collaboration and proactiveness in areas such as (1) more equitable access to smart healthcare, (2) more access to intelligent education, and (4) faster responses to emergencies.

Refers to the internet saturated in misinformation and AI “garbage.” According to MIT Technology Review, large language models are trained on data sets built by scraping the internet for text. This text includes silly, false, and malicious things humans have written online. The finished AI models regurgitate this content as fact, and their output is spread everywhere online. Tech companies scrape the internet again, scooping up AI-written text used to train bigger, more convincing models, which humans can use to generate even more nonsense before it is scraped again and again—continuing the cycle of “internet poisoning.” AI feeding on itself and producing increasingly polluted output also extends to images, resulting in the internet forever contaminated with images made by AI.

See: Digital Decay | Learn & Work Ecosystem Library

Four higher education interstate compacts in the United States facilitate cooperation among their member states to address common challenges, leverage resources, and improve educational opportunities for students within their respective regions. The compacts are nonpartisan, non-profit organizations that were established in 1948, Southern Regional Education Board (SREB); 1953, Western Interstate Commission for Higher Education (WICHE); 1955, New England Board of Higher Education (NEBHE); and 1991, Midwestern Higher Education Compact (MHEC). Established by either Congress or by agreements among the states themselves, the compacts together represent 47 states and territories and 6 state affiliate partners.

According to the AACRAO Higher Ed Glossary, a synchronized transcript of training, experience, and education acquired during military service in the United States. The JST is available to current and former military-service members in hard copy or an online delivery format.

The ways in which knowledge—scientific, social, and cultural—is produced has undergone significant and fundamental changes in the last century.

Knowledge production is often defined in a higher education context. It typically refers to related activities in a higher education institution, research center, or enterprise that is engaged with producing new knowledge. The term refers broadly to basic research as well as the more applied type of research associated especially with industry.

As described by the International Encyclopedia of Higher Education Systems, knowledge production has become a central task in modern universities. This represents a shift from the university’s origins in medieval Europe, which informed what became known in Europe and America as a “liberal education” —historically widely viewed as the purpose of universities, which in turn was closely associated with the supply of individuals prepared with the skills to meet the needs of the state, commerce, and the church.

There is growing awareness that knowledge production is moving beyond the main purview of higher education institutions, as industry invests more in research and development, driven in large part by rapidly accelerating science and technology developments.

Large Language Models (LLMs) are a class of artificial intelligence (AI) systems that consist of very large and diverse collections of text (and often code, images, and other data) that recognize patterns in language and generate human-like responses. LLMs are designed as general-purpose models capable of performing a wide range of tasks, such as summarizing, drafting, translating, reasoning, and answering questions across many domains.

LLMs do not “know” facts in a human sense; rather, they predict responses based on statistical patterns learned during training. Examples of LLMs include ChatGPT (OpenAI), Gemini (Google), LLaMA (Meta AI), Bard (Google AI), and Claude (Anthropic).

Depending on how they are deployed, LLMs may be further refined through techniques such as fine-tuning, feedback loops, or reinforcement learning, often with governance controls that limit how user interactions are retained or used.

In the context of the learn-and-work ecosystem, LLMs are useful for their breadth, flexibility, and ability to provide information and support exploration by users. Examples of use include career navigation, content generation, and knowledge synthesis. Their general-purpose design results in varying levels of accuracy, consistency, and explainability. For this reason, human oversight and organizational guardrails are needed to work effectively with LLMs.

See Glossary: Closed-Circuit AI Models (Domain-Specific or Enterprise AI Models) | Learn & Work Ecosystem Library

See Glossary: Opacity (in Artificial Intelligence) | Opaque AI | Learn & Work Ecosystem Library

See Topic Brief: AI Architectures in the Workplace: Large Language Models & Closed-Circuit AI Systems | Learn & Work Ecosystem Library

Launched in Fall 2022, the Learn & Work Ecosystem Library collects, curates, and coordinates digital resources to help users understand the nation’s complex ecosystem of education, training, employment, and work. Grounded in a commitment to open access and community stewardship, the Library invites users to suggest new resources and contribute to quality assurance by recommending edits. It was developed in collaboration with researchers from George Washington University’s Program on Skills, Credentials & Workforce Policy. The goal: to create a living, continually updated hub of knowledge—built by the community, for the community.

The Library functions as an information aggregator, organizing a wide array of resources—such as glossary terms, innovation initiatives, special topic reports, organizational profiles, and an archive of key documents and websites—designed to support a diverse set of ecosystem stakeholders. Key features include:

• Different types of content: Glossary, Key Initiatives, Topic Reports, Organizations, Library Lens reports, Index, Archive
• Time-stamping to indicate recency of content
• Digital accessibility compliance for disabled users
• Content available in multiple languages (12)
• Search options: key word, filter by stakeholders and types of content, AI chatbot for natural language queries/synthesis
• “About” section: describes the Why, How, and What around the Library, staff team, national advisory board, commitment to open access (Creative Commons License)
• “Top 5” monthly searches
• Links to external websites
• Newsroom for articles, blogs, reports
• Graphic depictions (maps) of relationships among searchable artifacts

In addition to its core collection, the Library partners with organizations to develop and host specialized resource collections. Through formal project agreements, the Library can ingest and curate partner-owned materials, organize them within a consistent structure, and maintain them over time. These services ensure high visibility, quality, and long-term value while relieving partners of the burden of ongoing content management.

See: Learn & Work Ecosystem Library | Learn & Work Ecosystem Library

Refers to the practice of giving preference in the college admission selection process to alumni relatives such as a parent, grandparent, or sibling. Many highly selective and prestigious higher education use legacy admissions as a factor in their selection process because they place a value on the connections and loyalty that come with having generations of families associated with their institution.

Alternative terms: legacy preferences, alumni connections

Refers to an educational approach in which learners acquire knowledge, skills, and competencies by designing, building, experimenting with, and improving physical or digital creations. Emerging from the broader Maker Movement of the mid-2000s—which celebrated do-it-yourself (DIY) innovation, digital fabrication, creativity, and collaborative problem solving—Maker Learning adapts these principles to educational settings.

Beginning in the early 2010s, schools, colleges, libraries, museums, and community organizations increasingly incorporated makerspaces and maker-centered experiences into teaching and learning. Over time, the emphasis shifted from providing access to tools and spaces to designing learning experiences that foster creativity, innovation, and the application of knowledge.

The term is rooted in the theories of experiential learning and constructionism. Maker Learning emphasizes learning through creating as opposed to relying primarily on lectures or memorization. This approach enables learners to investigate authentic problems, prototype solutions, test ideas, learn from failure, and continuously improve their work through reflection and iteration. Learning may occur in classrooms, makerspaces, fabrication laboratories (Fab Labs), libraries, community organizations, workplaces, or virtual environments using tools that range from craft materials to robotics, coding platforms, artificial intelligence, and advanced digital fabrication technologies.

Maker Learning is often associated with STEM and STEAM education, but it has expanded into many other disciplines, including workforce education, entrepreneurship, healthcare, the humanities, and lifelong learning.

The growing interest in Maker Learning reflects broader changes in education and work, including increased emphasis on competency-based learning, innovation, entrepreneurship, human-centered uses of artificial intelligence, and preparing learners to adapt to rapidly changing technologies and careers. Maker Learning has become an important instructional approach for preparing learners to succeed in an innovation-driven economy where the ability to create, apply, and continuously develop knowledge is increasingly viewed as being just as valuable as acquiring it.

According to the AACRAO Higher Ed Glossary, secondary-education (high school) alternative to a traditional transcript. Currently in use only by private institutions. Does not include standard letter grades but assigns mastery credits.

MCP stands for Meaning–Context–Probability, a concept that explains how information gains meaning when it’s used by people or by artificial intelligence (AI) systems. It helps us understand that information by itself is not enough—its value depends on how it is interpreted and used.

• Meaning refers to what the information represents.
• Context is the situation or setting in which the information is used.
• Probability reflects how likely it is that the information is relevant or accurate in that situation.

AI systems—and people—use all three parts together.  In today’s AI-driven world, understanding MCP helps users recognize that information is dynamic. The same data can have different meanings in different contexts—such as a “certificate” in higher education versus a “certificate” in cybersecurity training. AI tools rely on MCP principles to interpret search requests, connect related information, and deliver relevant results. For example, the Learn & Work Ecosystem Library Assistant bot (search tool) interprets a query based on what you typed (meaning), what part of the Library you’re searching (context), and what information is most likely to answer your question (probability). By keeping meaning, context, and probability in balance, the Library—and other AI-powered information systems—can help users find information that is not only accurate but also useful and understandable for their purpose.

Tips for Library users to apply MCP when searching with the AI Assistant bot at our site:

• Be clear about what you want to know. Example: Instead of “Tell me about credentials,” try “What are examples of stackable credentials in healthcare?” [meaning]
• Add where or how you’ll use the information. Example: “Explain credential transparency policies in community colleges.”  [context]
• Include what matters most to you—such as “most recent,” “U.S. examples,” or “employer-focused initiatives.”   [probability]

Related Terms

• Artificial Intelligence (AI) – the simulation of human intelligence in machines that can process information, learn, and make decisions.
• Digital Literacy – the ability to use digital tools and technologies to find, create, and communicate information.
• Information Architecture – the structured design of how information is organized, connected, and retrieved in a system.
• Information Literacy – the ability to locate, evaluate, and use information appropriately and responsibly.
• Metadata – data that describes other data to help systems and users find, understand, and use information effectively.

Refers to a rising number higher education institutions with very large enrollments and which commonly focus on flexibility, affordability, accessibility, and teaching and learning. Many of the mega-universities are optimized for working adults and have developed flexible online platforms for learning, often enhanced with artificial intelligence. Examples:

• Western Governor’s University (WGU): With over 156,000 students, its competency-based education model enables students to demonstrate skill acquisition and complete courses at their own pace, instead of within fixed credit hours. WGU has developed the Aera Decision Cloud platform to enable AI to predict student outcomes and recommend interventions to faculty.
• Arizona State University (ASU): With more than 150,000 students, it has expanded its online platform, ASU Digital Immersion, and with its partnership with OpenAI has launched an AI Innovation Challenge for faculty and staff to promote AI solutions across in all areas on campus.
• Southern New Hampshire University (SNHU): With approximately 158,000 students, its AI-powered chatbot “Penny” is designed to advise students on academics, finances, and wellness and has already increased student retention and performance.

Student experiences often differ in a mega-university from a traditional university. Students at a mega-university may never meet their professors because they’re taking asynchronous courses that provide little if any interaction with faculty. Students may also have few, if any, opportunities for interactive experiences on campus.

The role of faculty members may also differ. Institutions like SNHU and WGU have unbundled the role of the faculty member, to instead provide a life coach who stays with the student throughout matriculation. Additional features at these universities include teaching coaches plus expertise in instruction within the faculty such as WGU’s faculty who are dedicated to assessment.

See Topic: The Rise of Mega-Universities | Learn & Work Ecosystem Library

Refers to awareness and management of one’s own thinking and learning processes. This is often described as “thinking about thinking.” It includes planning how to approach a task, monitoring your understanding and progress, recognizing uncertainty or gaps in knowledge, evaluating results, and changing strategies when necessary.

Metacognition is particularly important when using AI because users need to consider what they understand themselves, where they need assistance, whether an AI response makes sense, and when additional information or verification is needed.

An acronym used by governments that stands for young people (typically ages 15 – 24) who are “Not in Education, Employment, or Training.”  Individuals categorized as NEET by governments are commonly unemployed, not enrolled in an education or vocational training program, not engaged in housework, and not seeking work.  The term is used to describe young people in order to exclude people in the retirement category (an older-age group).

The term emerged in the late 1990s in the United Kingdom and is now widely used among many countries including the United States.

Related Terms linked to unemployment: anti-unemployment, displaced, frictional unemployment, idle (not active or working)

Refer to the practice of companies agreeing not to recruit or hire each other’s employees.  No-poaching agreements may unfairly treat low-wage workers by keeping them locked in low-paying jobs without opportunities for advancement. These agreements are illegal under federal and state antitrust law. Employers that enter into these agreements may face civil and criminal penalties.

No-poach agreements also pertain to higher education institutions. They may face lawsuits for “no-poach” agreements that restrict faculty hiring across institutions.

See: Antitrust Lawsuits in Higher Education and Businesses | Learn & Work Ecosystem Library

Non-credit education includes any course or program that did not go through the process to be for-credit at a community college or university. They typically include personal enrichment classes, customized training for employers, English as a second language classes, and adult basic education. Many higher education institutions develop noncredit to credit bridge pathways to enable learners to earn credit for learning acquired through noncredit courses and programs.

Opacity (in artificial intelligence), often referred to as opaque AI, describes the degree to which an AI system’s internal processes and decision-making logic are difficult or impossible for humans to understand, interpret, or explain. An AI system is considered opaque when it produces outputs or recommendations without providing a clear, human-readable account of how those outcomes were generated. Opacity does not imply secrecy, error, or malicious intent. Rather, it reflects a structural limitation in interpretability, where the internal workings of the model are not easily transparent to users, decision-makers, or affected individuals.

AI opacity commonly arises in machine-learning and deep-learning systems that rely on large volumes of data, complex statistical relationships, and probabilistic modeling rather than explicit, rule-based instructions. In such systems, even developers may be unable to trace a specific output back to a sequence of logical steps.

In learning, credentialing, and workforce contexts, AI opacity raises important concerns related to trust, accountability, fairness, and governance, particularly when AI systems influence high-stakes decisions such as college admissions, employer hiring, assessment of learning, or credential recognition.

See Glossary Term:  AI Literacy vs. Adversarial Literacy | Learn & Work Ecosystem Library

See Glossary Term: Information Literacy | Learn & Work Ecosystem Library

See Topic Brief: Converging Terms:  Digital Literacy & Information Literacy | Learn & Work Ecosystem Library

Refers to a planning and risk assessment exercise in which a group imagines that a project, initiative, policy, or strategy has failed in the future and then works backward to identify what may have caused the failure. Unlike a traditional “postmortem,” which analyzes problems after something has gone wrong, a premortem is conducted before implementation begins. The goal is to identify hidden risks, flawed assumptions, unintended consequences, operational weaknesses, or overlooked challenges early enough to improve decision-making and reduce the likelihood of failure. The approach is especially valuable in periods of rapid change and uncertainty because it encourages people to think critically about vulnerabilities that may otherwise be ignored during optimistic planning processes. The growing use of premortems reflects a broader shift toward proactive planning approaches designed to help organizations prepare for uncertainty, complexity, and rapid technological change.

Premortems are increasingly used in strategic planning, project management, higher education transformation initiatives, technology implementation, AI governance and deployment, public policy, healthcare, business innovation, cybersecurity and risk management, and organizational change efforts

A typical premortem exercise asks participants to imagine: “It is two years from now, and this initiative failed. What happened?” Participants then identify possible causes such as unrealistic timelines, lack of stakeholder support, poor communication, insufficient training, ethical concerns, financial problems, technology limitations, data quality issues, regulatory barriers, and unintended impacts on workers or learners

Premortems are increasingly being used in discussions about artificial intelligence as organizations try to anticipate risks before deploying AI systems at scale. Examples:

• A college implementing AI advising systems may conduct a premortem to identify possible student privacy concerns, inaccurate recommendations, or unequal impacts on different student populations.
• A employer adopting AI hiring tools may use a premortem to examine risks related to bias, transparency, workforce resistance, or legal compliance.
• A government agency may conduct an AI premortem to explore cybersecurity vulnerabilities, misinformation risks, or public trust issues before launching an AI-enabled service.

Some organizations use AI tools during premortem exercises themselves. In these cases, AI systems help generate possible failure scenarios, identify overlooked risks, simulate stakeholder reactions, summarize patterns from prior project failures, or support scenario-planning activities. However, human judgment remains important because AI-generated risks may be incomplete, inaccurate, or lack organizational context.

Prison education refers to educational programs provided to individuals incarcerated in correctional facilities. These programs aim to equip those who are justice-impacted with knowledge, skills, and qualifications that can help them during their incarceration and after their release. Prison education is seen as a critical component of rehabilitation and a means to support the successful re-entry of formerly incarcerated individuals back into society.

Refers to the process that occurs when an institution of higher education sunsets an academic program. It typically involves removing the program from the higher education institution catalog, from the Student Information System (SIS), and from degree-audit systems. The institution then informs accreditors and the U.S. Department of Education.

Refers to digital technologies that are designed to strengthen human relationships, collaboration, trust, empathy, social support, and community. Prosocial technologies help people connect with, learn from, and support one another while using technology to enhance human capabilities and relationships. In education and workforce development, prosocial technologies—including AI-powered tools—can help learners and workers identify mentors, advisors, instructors, peers, employers, and professional networks; facilitate meaningful introductions and collaboration; encourage help-seeking behaviors; and strengthen the social connections that contribute to learning, career development, and personal success. The goal is not to substitute for human relationships, but to make them more accessible, effective, and sustainable.

An organization-level disclosure that clarifies what role AI did—and did not—play in producing knowledge, credentials, or guidance. These statements support trust, reproducibility, and informed interpretation while acknowledging the growing normalization of AI-assisted work.  These statements are increasingly required by publishers, journals, educators, funders, researchers, workforce organizations, and other oversight bodies to ensure transparency around AI-assisted creation, analysis, and decision-making. These statements commonly address:

• Scope of Use – Whether and how GenAI is used in content creation, editing, analysis, coding, translation, or administrative tasks; and whether use is optional, limited, or embedded in workflows.
• Human Oversight and Accountability – Describes review or approval processes and clarifies that humans retain responsibility for accuracy, interpretation, and final decisions.
• Data and Privacy Protections – Whether proprietary, personal, or sensitive data are excluded from GenAI tools Safeguards to prevent data leakage or unauthorized reuse.
• Validation and Quality Control – Methods used to check accuracy, bias, hallucinations, or misuse of information. Ongoing monitoring or auditing practices, where applicable.
• Ethical and Policy Alignment – Alignment with organizational policies, publisher guidelines, or professional standards; and disclosure of any restrictions imposed by funders, regulators, or partners.
• Transparency to Readers or Users – How AI use is communicated to audiences (e.g., footnotes, acknowledgments, methodology sections).

A relational map visually presents connections among entities within an ecosystem, such as organizations and initiatives. Relational maps help users understand the overall structure or domain of an area of interest.

In 2024, the Learn-& Work Ecosystem Library initiated relational maps to complement narrative descriptions of key searchable artifacts.  See examples in prototype maps. Maps are developed by integrating manual data tagging with inferred AI-driven relations. This work includes a unique collaboration with ChatGPT’s API under an open licensing agreement that allows the Library to train and continuously refine the AI model.

Related Terms: 

• Concept Map: Diagram that shows the relationships among ideas to help users understand how ideas are connected. Concept maps are generally composed of two elements: concepts (usually represented by circles, ovals, or boxes and are called nodes); and relationships (usually represented by arrows that connect the concepts; the arrows often include a connecting word or verb and these arrows are called cross-links.  There are four types of common concept maps: (1) spider maps, (2) flowcharts, (2) hierarchy maps, and (3) system maps.
• Mind Map: Diagram that shows the relationships among ideas to help users better understand, remember, and communicate information. Mind maps generally organize information into a hierarchy, showing relationships among pieces of the whole. A central concept or idea is usually placed in the middle of a spider diagram, with associated concepts/ideas that are connected branching out from the center (key words are called nodes).

Remedial education (aka developmental education) is required instruction and support for students who are assessed by their institution of choice as being academically underprepared for postsecondary education. The intent of is to educate students in the skills required to complete gateway courses, and enter and complete a program of study. Remediation at the postsecondary level is delivered at both community college and university campuses although some states have established policy to limit public university provision of remedial education. The bulk of remedial courses focus on advancing underprepared students’ literacy (English and reading) skills or math skills. Students are often placed into remedial courses through placement tests such as the ACT, ACCUPLACER, or COMPASS assessments. Typically, each college or university sets its own score thresholds for determining whether a student must enroll in remedial courses. Some states are moving toward a uniform standard for remedial placement cut scores.

Refer to a series of AI models that teach robots to complete basic tasks in environments they have never been trained for, without additional training or fine-tuning. RUMs allow machines to complete five different tasks: opening doors and drawers, and picking up tissues, bags, and cylindrical objects in unfamiliar environments. This is a significant advance since researchers typically need to train robots on new data for each new environment they encounter — often a time-consuming, expensive process.

As defined by the Ascendium Education Group, single mothers are women who are the primary caregivers and financial providers for their children without a co-parent or partner. Ascendium supports a grant program focused on:

• Improving educational attainment and career prospects for single mothers by increasing degree and other credential completion rates at participating community colleges. The initiative aims for a 30% increase in attainment rates by 2024, targeting more than 6,000 single mothers.
• Creating scalable models to design replicable and scalable solutions that other postsecondary institutions can adopt to support single mothers across the U.S.

See: Single Moms Success Design Challenge – Ascendium Education Group & Education Design Lab | Learn & Work Ecosystem Library (learnworkecosystemlibrary.com)

Refers to the process in which a state funds for variation in inputs across higher education institutions and enrollment changes annually. States calculate appropriations using a formula that accounts for specific inputs (e.g., number and characteristics of students enrolled, the level or field of study). States often codify allocation formulas through legislation, so legislators and governing boards have fewer opportunities to intervene.

Refers to a learner temporarily withdrawing from enrollment at a college or university, or choosing not to re-enroll in an ongoing degree program, often to pursue another activity or due to competing obligations. Stop-out is distinct from the concept of drop-out, because the pursuit of education is delayed rather than abandoned.

Refers to information that is artificially created rather than collected from real people or events. It is generated by computer programs to mimic the patterns and characteristics of real data without exposing anyone’s personal details. For example, instead of using actual student records to test a new education tool, developers might use synthetic data that looks and behaves like real student records but does not belong to any actual person. This allows organizations to test, train, and improve systems while protecting privacy and reducing risks.

The term simulated data is sometimes used in connection with synthetic data, especially when the information is produced through a computer simulation. In practice, all simulated data is synthetic, but not all synthetic data comes from simulations—some is generated by machine learning models, statistical methods, or rule-based systems.

Refers to the circumstance in which a higher education institution must provide completion opportunities for impacted learners when the institution discontinues an academic program or closes or ceases operations of an academic program.

Refers to an institution of higher education in the United States that is established and operated by a federally recognized American Indian tribe or Alaska Native community. These colleges primarily serve Native American populations and often focus on preserving and promoting indigenous cultures. There are 35 Tribal colleges and universities in the U.S. They are located in 14 states, primarily in areas with significant Native American populations.

Refers to the use of Short Message Service (SMS) texting or mobile messaging platforms that enable real-time, interactive communication between students and educational institutions. Unlike one-way broadcast alerts or notifications, two-way texting allows learners to respond to messages, ask questions, request assistance, and receive personalized or automated support through mobile messaging. The approach combines mobile communication, behavioral science, and conversational technologies—often supported by artificial intelligence (AI)—to deliver timely, accessible, and personalized guidance throughout the learner lifecycle. The approach is increasingly aligned with broader shifts toward personalized learner support, data-informed advising, and mobile-first service delivery models.

Two-way educational texting emerged in the late 2000s with “nudge” interventions designed to improve college enrollment and student success outcomes. Early randomized trials demonstrated that targeted text reminders could significantly increase college enrollment and financial aid completion behaviors.

During the 2010s, dedicated education messaging platforms expanded, particularly within enrollment management and advising operations. Adoption accelerated substantially during the COVID-19 pandemic as educational institutions sought scalable mobile communication tools capable of supporting remote student services. Because texting is widely accessible across demographic groups and does not require specialized applications or broadband connectivity, it is frequently used to improve communication equity and responsiveness.

Today, conversational texting is widely used across K-12, postsecondary, and workforce training environments. Community colleges, online learning providers, and adult education programs have been among the most active adopters due to the need to support diverse and often nontraditional student populations.

Two-way texting platforms typically include:

• Real-time conversational communication between students and staff or automated virtual assistants
• Personalized and targeted messaging based on student milestones, risk indicators, or service needs
• Immediate responses to student questions or requests for assistance
• Integration with advising, financial aid, tutoring, enrollment management, and student success systems
• Availability outside traditional business hours through automated or AI-supported messaging
• Data tracking and analytics to support student success interventions and institutional decision-making

Evidence from research studies, pilot programs, and institutional implementation reports suggests that two-way texting can:

• Improve student engagement and responsiveness to institutional communication
• Increase completion of enrollment and financial aid processes
• Strengthen advising and tutoring participation
• Support persistence, particularly among part-time and first-generation learners
• Expand access to support services outside standard operating hours
• Foster stronger student-institution connection and sense of belonging

Examples of platforms supporting two-way texting and conversational messaging in education include:

• Remind — K-12 communication platform supporting teacher-student-family messaging and classroom communication
• Ocelot — AI-enabled conversational student engagement and support platform used primarily in higher education
• Mongoose (Cadence) — Higher education conversational texting and student engagement platform used for recruitment, advising, and retention communication
• Modern Campus Message — Student lifecycle communication and engagement platform integrated with student information and success systems
• Voxer — Push-to-talk and messaging platform used in some advising, coaching, and training environments
• Persistence Plus — Behavioral messaging and coaching platform focused on student motivation, nudging, and persistence support

Refers to individuals who reside in the United States without legal documentation or authorization. They may have entered the country illegally or overstayed their visas. Despite their undocumented status, many have been raised and educated in the U.S. during elementary and secondary school years, and often face significant barriers to accessing higher education and other opportunities. Undocumented students are estimated to comprise a small percentage of the total student population—some studies suggesting less than 1% (an estimated 400,000 students including DACA recipients enrolled in higher education.)

See: DACA

Refers to the variation in how different people interact with artificial intelligence systems. Even when people use the same AI tool, their approaches, expectations, and interpretations may differ significantly. This variability influences how effective the AI is in practice and can affect whether systems succeed, stall, or produce inconsistent results in real-world settings. For example, users may:

• Ask questions in different ways.
• Use different levels of detail when providing instructions or context.
• Have different expectations of what AI systems can do.
• Interpret AI responses differently.
• Incorporate AI outputs into their work in different ways.

Because of this variation, the same AI system can produce very different results depending on the user.

In early testing and research on AI systems, differences in how people used these tools were sometimes treated as “noise” in the data—something to average out in order to measure system performance. More recent research suggests this variability is an important signal. Studying how people actually use AI can reveal why systems perform well for some users but not others, and why certain deployments succeed while others encounter unexpected risks.

Several related practices and terms include:

• Prompt engineering focuses on designing prompts—questions, instructions, templates, or workflows—that guide AI systems toward more reliable outputs.
• Prompt templates structure how users interact with AI systems.
• Prompt chaining sequences multiple prompts to generate more complex results, improved human–AI interface design, and efforts to strengthen AI literacy so users better understand how to interact effectively with AI tools.

An informal, emerging term that describes a way of creating digital tools, software, content, or workflows using artificial intelligence (AI) by focusing on what an individual wants to achieve rather than how it is technically built. Instead of writing detailed code or step-by-step instructions, users describe their goals, outcomes, or intent in natural language and rely on AI tools to generate and refine the underlying technical components through iteration and feedback.

The term originated in software development communities but increasingly reflects a broader shift across learning and work. AI tools now allow people without advanced technical training to build things that once required specialized expertise. As a result, capabilities previously considered advanced technical skills are becoming accessible to a much wider group of users.

In the learn-and-work ecosystem, vibe coding is best understood as a signal of changing skill expectations and work practices rather than a formal methodology or best practice. It highlights opportunities created by AI-enabled tools but also the ongoing need for human responsibility. Although AI can write code that works through vibe coding, it cannot guarantee the code is correct, safe, or appropriate. Human oversight remains essential, and the type of human oversight may change—from technical execution to judgment, validation, and responsibility.

The open internet ecosystem relies on web crawlers—automated bots that systematically browse millions of websites to collect various forms of data, including text, tables, images, audio, and video. Web-crawled data serve multiple purposes, such as:

• Powering search engines
• Tracking product and service prices across companies
• Informing digital platforms that aggregate information in industries such as travel, hospitality, job matching, and credentialing (education and training providers)
• Enhancing web security monitoring
• Supporting historical archiving
• Facilitating investigative research by government agencies, policy organizations, and think tanks
• Training artificial intelligence (AI) systems

Estimates suggest that crawler traffic accounts for nearly half of all internet activity and is poised to surpass human-driven traffic. However, the rapid expansion of AI-powered crawlers threatens the transparency and accessibility of the internet. Many websites risk displacement as AI crawlers increasingly dominate web traffic.

In response, website owners are implementing protective measures such as logins, paywalls, and anti-crawling technologies that detect, restrict, block, or charge fees for nonhuman traffic. These actions are fragmenting the internet, creating areas where AI crawlers have limited, slower, or no access—ultimately reducing information availability for human users and reshaping the concept of an “open” internet.

Large tech companies can afford to license extensive datasets and develop advanced AI web crawlers capable of circumventing these restrictions. In contrast, smaller content creators—such as visual artists, YouTube creators, and independent bloggers—may choose to hide their work behind logins and paywalls or remove it from the internet altogether. This shift risks concentrating control over the information ecosystem in the hands of AI developers and large data publishers.

To preserve an open internet, advocates will likely turn to laws, policies, and technical infrastructure aimed at protecting non-commercial and noncompetitive uses of web data.

Refer to a higher education institution approved by the U.S. Department of Education that meets the federal regulatory requirements to integrate work experience into their academic programs, fostering a unique and holistic approach to education. Guided by statute, Work Colleges must meet the requirement that all their resident students participate in a comprehensive work-learning-service program for all years of enrollment.

There is rapid growth of virtual reality (VR), augmented reality (AR), mixed reality (MR), and extended reality (XR) technologies and their applications.

• VR is a computer-generated simulation of a virtual, interactive, immersive, three-dimensional (3D) environment or image that can be interacted with by using specific equipment.
• AR is a technology used to create virtual objects (text, images and sounds) being superimposed onto a three-dimensional real-world environment using projection, optical, or video see-through devices.
• MR comprises AR and VR.
• XR is an umbrella term including VR, AR, MR, and virtual interactive environments.  In general, XR simulates spatial environments under controlled conditions, to enable interaction, modification, and isolation of specific variables, objects, and scenes in a time- and cost-effective manner. The main applications of the XR combination with artificial intelligence (AI) are autonomous cars, robotics, military, medical training, cancer diagnosis, entertainment, gaming applications, advanced visualization methods, smart homes, affective computing, and driver education and training.