The Current State of Play: AI in Higher Education and the Road Ahead

The age of AI offers an opportunity to rethink learning from the ground up. Paul LeBlanc and his colleagues are exploring a human-centered approach that could "leapfrog personalized learning" and enable a new level of precision learning.

For decades, higher education has quietly shifted its mission from human development to workforce development. The first five articles in our series largely focused on the future: a vision of what is possible with N-of-1 learning, the opportunity for faculty to focus more on relationships and human development, and ways to rethink curriculum and content, among other possibilities. We are optimistic about the future of learning. We believe AI can improve, democratize, and extend education to millions around the globe—and help higher education reclaim its broader purpose of developing humans.

Higher education as a sector is still the early stages in this transition, but it is no ordinary one. The speed and scope of change, visible across all sectors of society, can trigger a kind of institutional paralysis that sets in when change moves too quickly and too widely to be managed incrementally. "Deer in the headlights" may be too strong a metaphor, but "deer suddenly alarmed and not knowing which direction to run" may be close. Point solutions and patchwork approaches might be suitable for improving the current ways of doing things, but they are not adequate for a paradigm shift. As a sector, higher education needs to move faster and plan bigger.

However, colleges and universities are highly complex systems and among the slowest to change of any societal sector. Higher education is attempting to address the AI revolution, but it is applying tactical fixes to a structural crisis, running pilots where strategies are needed, and issuing AI-use policies instead of debating learning philosophies. These choices are not being made because people are not paying attention; if anything, it seems as if AI is the only topic being discussed at conference after conference. Leaders are awake and even alarmed. But wakefulness is not the same as movement, and alarm is not the same as action.

Ten conditions define the current moment, surfacing questions that leaders must grapple with to move their institutions forward more quickly. Some are crises hiding in plain sight, existing before AI came calling, and are now fully surfaced and unavoidable. Others are genuine opportunities kept just out of reach due to caution across the sector. All of them are interconnected, and none will resolve on their own.

1. The Data Foundation Is Broken, and It Is Not Just a Technology Problem

Ask any CIO or chief AI officer (an emerging role) what stands between their institution and meaningful AI deployment, and the answers are almost always the same: data, siloed systems, legacy infrastructure, and governance structures that are built for another era. Security teams are already buckling under AI-powered cyberattacks. The problem is arguably more organizational and political than technical. Data is not siloed by accident; it becomes siloed because the components of the tech stack don't play well with each other and, more insidiously, because departments treat information as a source of institutional power or control. Governance stalls not because people do not understand the technology, but because those in charge cannot agree on who is responsible for what.

Getting data architecture right (including governance and security, among other things) is likely the single greatest impediment to realizing the potential of AI to give higher education institutions a holistic understanding of their students, enable precision learning, improve insight and outcomes, and support agentic solutions that can reduce costs, improve the student experiences, and so much more. It is no surprise that this issue ranked as the number one issue in the 2025 EDUCAUSE Top 10.^Footnote1

The cybersecurity dimension of the problem deserves particular attention. The sophistication of AI-powered hacking attempts is outpacing institutional defenses, and chief information security officers are operating under considerable stress. Next-generation models such as Mythos and Spud are expected to raise the data security stakes even further, already causing panic among governments and financial institutions. As cyberattacks become more sophisticated (and frequent), colleges and universities must improve their data architecture and secure their data before pursuing more ambitious AI initiatives.

At the same time, security-minded IT departments can inadvertently slow the deployment and adoption of AI tools. Here's a simple test: if it takes weeks to get approval for an AI tool that the rest of the world is using already, your institution is no better off than a wounded antelope on the Serengeti—not a good position to be in when so many lions are afoot. Data can no longer be relegated solely to the IT department; it is a leadership, governance, and strategic matter worthy of the attention of the board and executive leadership.

2. Assessment Was Already Broken. AI Just Made It Undeniable.

For decades, higher education has measured learning through proxies—artifacts such as essays, exams, and problem sets—all of which are imperfect signals of actual learning and competence. These assessment mechanisms persisted not because they worked especially well, but because they scaled. A professor with two hundred students could not conduct ten-minute oral examinations with each individual. The lecture-quiz-grade model was a triumph of logistics over pedagogy and serves as a reminder that education in 2026 is rooted in an industrial model that prioritized efficiency, scale, and predictability over actual learning. It also framed assessment as judgment, motivating students to achieve a certain grade rather than master specific knowledge or skills.

AI has not broken assessment; it has exposed the deep inadequacies that were already there. When a student can produce a polished essay in minutes, the essay no longer reveals anything reliable about what that student knows or can do.^Footnote2 Many faculty responded by banning AI or going back to handwritten exams. These reactions are understandable, but they mistake the symptom for the disease. The deeper problem is that assessment was designed to produce grades, not to measure learning, and the two have never been the same thing.

More importantly, AI makes genuine assessment of student learning newly possible if we, as educators, are willing to rethink our assessment practices. The Socratic oral examination, considered the gold standard for revealing what a student actually understands, can now be conducted at scale. Dialogic assessment, where what matters is not the artifact but the conversation about how it was produced, what choices were made, and why, is now practical and scalable with well-designed learning agents. Every experienced teacher knows that a ten-minute conversation can reveal whether a student "knows their stuff." Other AI-based technologies can support performance-based assessment through the analysis of student activity as they engage in tasks such as leading a team discussion, practicing the cello, preparing a dish, or delivering a presentation.

The wall between instruction and assessment was never pedagogical; it was a logistical compromise educators retrofitted with justifications about rigor and objectivity. AI collapses that wall. Assessment can instead be longitudinal, embedded, and omnipresent, with evidence of learning gathered continuously in context across the arc of a learner's work, rather than through high-stakes performance events. That collapse changes the psychological contract: no single moment carries disproportionate weight, and the work itself generates the evidence. This approach is more equitable by design: continuous embedded evidence captures growth and process rather than privileging those who perform best on demand in a standardized format. However, the collapse is only liberating if what fills the merged space is designed to support learning rather than detect failure. Ubiquitous surveillance would deliver the worst of both worlds. Motivation over detection provides the assessment system that higher education needs—and should have built.

The assessment crisis is real, but it is also an invitation. Institutions that seize it will produce graduates who can demonstrate not just what they know but what they can do with that knowledge.

3. Policy Without Philosophy Is Just Noise

Although many colleges and universities have developed some form of AI use policy, not all have articulated what they actually believe about learning and technology use. The result is a proliferation of guidance documents that signal institutional awareness but provide little support for decision-making on genuinely complex trade-offs. This ambiguity has become a source of enormous frustration for students, most of whom want to do the right thing.

To the extent that higher education institutions now have AI policies in place, one could argue that the sector has not solved the problem so much as documented it. What is needed is a working theory of learning in the age of AI: What cognitive work should students do themselves? What work should AI augment? And what does genuine competence look like within each discipline when the tools have changed? What obligations do institutions have to graduates regarding the tools they should have mastered and the AI capabilities employers will expect? These are fundamentally epistemological and philosophical questions before they are policy questions, yet many institutions have moved quickly to the latter without fully considering the former. In some cases, they have left the question to individuals, which is close to having no policy at all.

We suggest a three-level scaffolded policy structure that begins with a foundation of institution-wide principles, such as ethical and responsible use, data management, acceptable tools, and transparency guidelines. The next level should include unit- or department-specific policies; for example, what is appropriate for the Financial Aid department, the School of Engineering, or the Student Affairs office. The top level should identify those decisions within university or departmental frameworks that are left to individuals.

4. Agentic AI Has Come for the Back Office

Administrative functions across higher education are already being reshaped by agentic AI, whether institutions have planned for it or not. Companies developing student advising systems, early alert platforms, enrollment management solutions, financial aid processing systems, and content development tools are rapidly integrating agents into their products. At the same time, new AI-first point solutions are emerging, and some institutions are developing their own agents. The transformation is not coming; it is already underway.

The two challenges institutions are least prepared for are also the most difficult. The first is workflow reinvention. Dropping an AI agent into an existing process is not enough; the process itself typically needs to be rethought from the ground up. The agent reveals how many steps in the current workflow exist not because they are necessary but because they were shaped by the limitations of pre-AI systems. Rethinking these workflows requires time, authority, and a willingness to challenge institutional habits—none of which are in abundant supply. Simply bolting AI onto existing processes misses the opportunity to improve them.

The second challenge is a staffing conversation that few in higher education are willing to have openly. One for-profit online provider recently replaced a twenty-person content development operation with two people and AI. Another AI-first solution provider promises a minimum 20 percent reduction in workforce if an institution wishes to take that step. New vendors provide "agentic workers" designed to take over entire areas of administrative operations. Resistance to such changes will almost certainly be fierce, but the overall direction is clear. Back-office staff will eventually shrink. The roles that survive will increasingly involve managing and assuring quality control of the agents' output rather than doing the work themselves.

5. The Value of Static Curriculum Is Collapsing

For centuries, the syllabus was an act of curation with genuine scarcity value. The professor had traveled the intellectual terrain; the student had not. The professor knew which texts mattered, which debates were central, and which sequence of encounters would produce understanding. That curation was worth something.

Today, the knowledge graph is doing much of that work, and it is doing it faster, more comprehensively, and more adaptively than any fixed syllabus can. The static, instructor-owned, closed curriculum will increasingly seem both limited and limiting. More important than the curriculum is the design of the agent that delivers it. This is one place where the role of the faculty remains critical, as they bring judgment about what matters and why, help learners navigate toward wisdom rather than mere information, and design the kinds of pedagogically informed encounters that develop genuine competence.

The role of the instructional designer is also being fundamentally reshaped. Where instructional designers once produced courses for human learners, they are increasingly designing for machine-mediated learning—for AI agents that will meet humans as they learn. The work has shifted to structuring the knowledge graphs that represent a domain, specifying the pedagogical priors that shape how an agent interprets a learner's state, designing assessment logic that distinguishes performance from completion, and tuning the feedback loops that enable the whole system to improve over time. Rather than delivering a course, the designer is now responsible for designing the architecture from which a "course of one" emerges for each learner.

The curriculum is no longer a fixed object. It is a dynamic system, and designing it well will require a new kind of collaboration between designers, faculty, and technology.

6. AI Is Now Superb at Knowledge Transfer

This condition makes faculty the most uncomfortable, so it is worth being precise about what it does and does not mean. AI does not replace the professor. It replaces the old "one-size-fits-all" model of knowledge delivery in which an expert broadcasts content to a room of students at varying readiness levels, most of whom retain only a fraction of what they hear, and few of whom ask the question they most need answered. That model has always been a poor use of the most expensive and irreplaceable resource in higher education: the time and attention of a human expert.

What AI now makes possible is a profound reallocation. A "genius" teaching assistant—nonjudgmental, inexhaustible, available at any hour, and capable of detecting when a student's mental model breaks down and offering a targeted correction—handles basic knowledge transfer. This frees the instructor to do what only humans can do well: build relationships that make students feel seen and valued, hold space for the ethical and existential questions that shape lives, and model the integration of knowledge with wisdom, judgment, and character. Research on developmental resilience consistently shows that the presence of one caring adult who takes a genuine interest in a young person can positively influence outcomes.^Footnote3 AI cannot provide that. Faculty can, if they are freed from cognitive labor that a machine can now perform more efficiently. In this new model, faculty become curators of community, caretakers of relationships, and scholarly role models who enrich students' understanding of their discipline.

In many respects, this is a more demanding role for faculty. The functions that remain irreducibly human aren't "soft"—they are the high-skill core of teaching: judgment formation (modeling expert decision-making at the frontier), identity apprenticeship (conferring membership in a discipline through being seen and taken seriously by someone who already belongs), productive struggle calibration (knowing when to intervene and when to wait), and the introduction of "unknown unknowns" (surfacing questions the learner—and the AI—didn't know to ask). Large language models (LLMs) can support aspects of introductory-level judgment formation and externalize reasoning patterns at scale, but they tend to model consensus judgment, rather than frontier judgment. They can't convey the practitioner's stake in the answer—the lived experience that turns reasoning into judgment. AI is fundamentally a known-unknowns machine; some of the most generative learning happens when someone in the room reframes the problem itself, and that requires humans who don't share the same blind spots.

Many faculty already do this work exceptionally well, particularly in apprenticeship-intensive disciplines and among master teachers across fields. However, that expertise is unevenly distributed and systematically under-resourced. Institutions have long rewarded the type of transfer-oriented teaching that AI just made cheaper, while undervaluing the work that actually forms practitioners, including the time the best teachers devote to students. At many institutions, tenure and promotion systems continue to reward knowledge work and not student-focused work. The task, then, is not to remediate faculty but to better support their work by making visible and learnable what skilled faculty already do—cognitive apprenticeship, fluency in signature pedagogies, formative assessment grounded in the learning sciences, and frameworks for productive struggle—and providing real training in AI as pedagogical infrastructure so faculty can use it to extend their capacity for the work that creates impact rather than compete with AI on knowledge transfer.

To achieve this goal, institutions must also rethink the underlying conditions: smaller cohorts and longer relationship arcs (calibration and apprenticeship are impossible at scale); promotion criteria that recognize this work as scholarly and valuable rather than mere service; protected time for the wide reading and intellectual drift that fuels the unknown-unknowns function, physical and social architecture that enables cross-pollination, and explicit acknowledgment that this is a system-level evolution, not a faculty performance problem. Without that infrastructure, faculty will rationally optimize for the transfer work AI just made cheaper, and the evolution becomes another unfunded mandate dressed up as innovation. The institutions that figure this out will reallocate real resources toward the functions only humans can perform. The ones that don't will end up with the worst of both worlds: AI layered onto legacy structures, and faculty unable to fully realize the irreducibly human work they are uniquely positioned to do.

Faculty members will need to embrace this shift in their roles. Some will be more comfortable with lecturing or "broadcast" pedagogy and less interested in forging stronger bonds with students. Until institutional reward and recognition structures align with these new expectations, progress is likely to be uneven.

Paul LeBlanc highlights a shift from personalized learning to AI-driven "N-of-1" precision learning, where instruction adapts in real time to each learner's needs and context, enabling individualized pathways to shared outcomes and redefining the classroom as a space for human connection.

7. Personalized Learning Is About to Be Leapfrogged by Precision Learning

The dream of personalized learning (tailored pacing, adaptive content, and individualized pathways) has been on the horizon in educational technology for two decades. In practice, it has arrived only in modest forms: adaptive problem sets, remediation loops, and basic differentiation. And it has consistently fallen short of the goal.

AI enables educators to leapfrog personalized learning (which, in many respects, is a form of population segmentation) and make something far more powerful possible: precision learning, in which the entire learning experience is engineered around the specific profile, goals, gaps, and context of a single learner. This paradigm shift moves higher education away from the aforementioned "one-size-fits-all" education model of the Industrial Age toward genuinely student-centered learning fine-tuned to each learner.

The potential gains from this shift are significant, particularly for learners who have historically been least well served by the standardized model, including first-generation students, working adults, those whose needs do not fit neatly into the conventional academic calendar, neurodivergent learners, and others. But precision learning also introduces a risk that deserves serious attention. When each student has a fully individualized experience, what do they hold in common? In a society already marked by rising loneliness and civic fragmentation, the inadvertent loss of shared intellectual community could be a serious cost.^Footnote4 Institutions that pioneer precision learning will need to be equally intentional about designing for connection as they are about designing for individualization. Values-based institutions offer one model of what this might look like, as students enrolled in religiously affiliated institutions, service academies, and Historically Black Colleges and Universities often master their discipline within strongly defined communities organized around shared values.

8. AI Allows Everyone in Higher Education to Better Know and Support Learners as Whole People

The convergence of learning analytics, AI advising systems, early alert platforms, and wellness monitoring has created something genuinely new: the technical capacity to understand a student not as a transcript but as a person. Which students are struggling academically and why? Which are showing signs of disengagement before they withdraw? Which may be experiencing mental health crises that their academic record does not yet reflect? The data to support learners holistically is increasingly available. What most institutions lack is the architecture to connect it, the governance to use it responsibly, and the human capacity to act on what it reveals.

There is a genuine tension here between support and surveillance that the sector has only begun to address. Students have a legitimate interest in privacy, but they also have a legitimate interest in being seen. Navigating that tension requires explicit conversations about consent, what data is collected and how it is used, and who benefits. Colleges and universities must address data challenges and deeper ethical, privacy, and relational questions. Getting these things right could improve persistence and graduation rates and result in happier, healthier students.

9. The Future of Work Is Unknown

Despite the difficulties it faces, the core value proposition of higher education remains durable: a degree opens doors, signals competence, and prepares graduates for careers that provide economic security and personal meaning. While higher education often positions itself as doing much more than career preparation, survey after survey shows that "getting a good job" remains the number one reason students go to college.^Footnote5 Preparing the next generation for the workforce is what the government primarily pays institutions to do, and it is what they increasingly hold colleges and universities accountable for in policies and regulations.^Footnote6 That reality was made abundantly clear in the 2026 U.S. Department of Education's gainful employment regulations, which now hold all post-secondary programs accountable for their graduates' earnings.^Footnote7 That mission is now genuinely in question, not so much because higher education has failed, but because the nature of work itself is destabilizing faster than any institution can keep pace with.

No one knows which jobs will exist in ten years, which skills will command a premium, or how the relationship between credentials and employment will be restructured by AI.^Footnote8 At a recent conference bringing together leaders in education and workforce development, sophisticated, thoughtful, and genuinely engaged students described deep anxiety about their economic futures despite their degrees. The enrollment surge in vocational and trade programs is the market sending a signal: students and families are seeking pathways that feel more stable amid economic and technical uncertainty. Much of that uncertainty is being shaped by rapid advances in AI—and what they mean for the future of work.^Footnote9 Colleges and universities that respond only by adding AI literacy courses to existing programs are underestimating how much the workforce landscape is changing and will continue to do so. The deeper challenge is to prepare graduates not for specific jobs that may not exist but for continuous adaptation, including the metacognitive agility to keep learning, navigate ambiguity, and work alongside systems that will continue to become more capable. Resilience and "thinking about thinking" will be more important than competence within a specific field.

If preparing students for specific jobs is no longer enough, higher education can reclaim its focus on developing the whole person, not just preparing students for the workforce. If undergraduate higher education is primarily about job preparation, it falls short of what most of us who work in it want it to be. Higher education should develop graduates with character, an ethical and moral compass, and a sense of civic responsibility. Students have been trained to think about their college or university education as a path to a successful career, but the best of them want so much more. They want an educational experience that sees them as individuals, makes them feel like they matter, and helps them dream bigger.

10. Almost No College or University Has a Real Strategy for This Moment

The last and perhaps most consequential condition is institutional: most higher education institutions remain in what we call play-and-pilot mode. They have AI task forces, policy working groups, faculty workshops, communities of interest, and a handful of experimental deployments. What they do not have is a governing theory of where they are going, a serious analysis of what they are willing to change and what they need to protect, or the organizational structure to execute transformation at the pace the moment requires.

The conventional explanation for this strategic vacuum points to the speed of technological change; it is moving too fast for institutions built for deliberation. That is true. . . and incomplete. The deeper issue is cultural. In fairness to higher education, many industries are struggling to keep up with the pace of AI advances. Higher education, however, moves even more slowly and is not built for the kind of transformational speed now underway. Getting institutional stakeholders to engage, rethink the work, and move faster may be the central challenge facing presidents and chancellors today, and that's saying a lot in such volatile times.

Meanwhile, the competitive landscape is shifting under everyone's feet. AI-native startups and large platform providers are building learning experiences that are more responsive and more adaptive than anything most traditional colleges and universities can offer. Corporate learning programs are expanding. Vocational enrollment is surging. Employers are demanding workers with AI skills, even as they remove the lower rungs of the organizational ladder, exacerbating an already difficult job market for recent graduates. The institutions that are most financially vulnerable—those that most need to reinvent—are also the ones with the least capacity to do so.

The culture problem is not insurmountable, but it requires leadership with both the courage to name it and the political skill to work within shared governance while still moving with purpose. Perhaps the best example we have seen of getting this work right is at the University of California San Diego (UCSD), where Chancellor Pradeep Khosla has created the infrastructure and support to encourage all stakeholders to be "builders," taking a let-ten-thousand-flowers-bloom approach to solving problems and reimagining them with AI. Eschewing a top-down approach, UCSD has provided tools, support, and sandboxes where anyone can experiment and build with AI, unafraid to break things or make mistakes. And when a small group of students inadvertently brought down the UCSD registration system with an AI app they built, the university hired them rather than punish them. Sure, the CIO and CISO may have a few more gray hairs, but UCSD is racing ahead of so many others in learning, building, and integrating AI across the enterprise. There are others, such as Western Governors University with its WGU 2.0 initiative, Syracuse University with its agentic deployment, and the UK's Ardent University, which is setting out to reinvent itself as an AI-first university. But at this point, these are exceptions.

Reclaiming the Purpose of Higher Education

Several conditions hover just outside this list but deserve mention in any honest accounting. The growing anti-AI sentiment among faculty, students, and the public is real and will likely intensify. This is not simply resistance to change; it is a legitimate expression of value conflicts over authenticity, human connection, and the role of technology in education. The risks of AI anthropomorphism, of students forming parasocial attachments to systems that are actively optimizing for engagement, represent a real and measurable risk on campuses already grappling with a mental health crisis.

Equity concerns are also deeply under-addressed. The gap between institutions that can afford to invest at scale (UCSD, for example, has a supercomputer and the resources to build its own LLM) and those that cannot is widening. Within institutions, the benefits of AI adoption are already unevenly distributed. At the same time, higher education's social contract with society, its claim to be a public good worthy of public trust and public investment, is fraying at precisely the moment it most needs to be renewed.

Taken together, these concerns might suggest that higher education is perilously close to failure or collapse in the age of AI. We believe the opposite can be the case. Institutions rooted in human relationships, committed to truth-seeking, and oriented toward the full development of persons play a central role. AI cannot manufacture the experience of mattering to another human being. It cannot model intellectual courage or ethical discernment. It cannot build the kind of community in which students discover who they are and what they believe.

These are not small things. They are, in fact, the things most worth doing. At their best, colleges and universities are not only preparing better workers but shaping individuals and strengthening society.

If AI is remaking the world—creating new problems and, perhaps, solving some of the most intractable ones—then the world will need what higher education institutions are uniquely positioned to provide: the deep expertise of faculty and researchers across disciplines. Technologists, philosophers, ethicists, theologians, psychologists, sociologists, and economists are needed now more than ever. Higher education still does science, research, and scholarship better than any other sector.

We believe colleges and universities are the best bet for getting it right. The institutions that will be most relevant in twenty years will be those that refuse to look away from today's challenges—those that can reinvent how they operate, serve students, and define their mission while remaining deeply committed to their core values: rigor, research, human flourishing, and the advancement of knowledge in service to society. AI can be seen as a threat or embraced as a powerful tool that helps higher education address its shortcomings and reclaim its central and critical role in society.

Authors' Note

The structure of the article was conceived collaboratively. Paul wrote the first draft and made revisions based on team feedback. We then asked Claude to analyze and critique the list, and Paul addressed some of its suggestions in the next draft. The team reviewed the article again, and Paul incorporated edits based on that feedback. Finally, we used Claude to identify some of the references, which we double-checked for accuracy.

1. Deborah Dent, James Frazee, Carrie Shumaker, and Tim Wrye, "2025 EDUCAUSE Top 10 #1: The Data-Empowered Institution," EDUCAUSE Review, October 23, 2024.Jump back to footnote 1 in the text. ↩
2. Chrysanthos Dellarocas, "AI Will Break Assessment Before It Fixed It," Inside Higher Ed,February 19, 2026.Jump back to footnote 2 in the text. ↩
3. Emmy E. Werner, "Vulnerable but Invincible: High-Risk Children from Birth to Adulthood," Acta Paediatrica Supplement 422 (July 1997): 103–105; Michael Rutter, "Psychosocial Resilience and Protective Mechanisms," American Journal of Orthopsychiatry 57, no. 3 (1987): 316–331.Jump back to footnote 3 in the text. ↩
4. Vivek H. Murthy, Our Epidemic of Loneliness and Isolation: The U.S. Surgeon General's Advisory on the Healing Effects of Social Connection and Community (U.S. Department of Health and Human Services, 2023).Jump back to footnote 4 in the text. ↩
5. Gallup, The State of Higher Education 2024, (Lumina Foundation, May 2024). Jump back to footnote 5 in the text. ↩
6. Gallup and the Strada Education Network,Why Higher Ed? Top Reasons U.S. Consumers Choose Their Educational Pathways (Gallup, 2018). Jump back to footnote 6 in the text. ↩
7. "2026 Gainful Employment," National Association of Student Financial Aid Administrators, January 30, 2026Jump back to footnote 7 in the text. ↩
8. Business Higher Education Forum, "New Report Highlights the Four Skill Sets Transforming Work," news release, December 1, 2022; Nik Dawson et al., How Skills Are Disrupting Work: The Transformational Power of Fast Growing, In-Demand Skills (The Burning Glass Institute and BHEF, December 2022).Jump back to footnote 8 in the text. ↩
9. National Student Clearinghouse, "Vocational-Focused Public 2-Years Lead Spring Enrollment Rise," NSC Blog, May 22, 2025.Jump back to footnote 9 in the text. ↩

---

Tanya Gamby is Vice President of AI Learner Development at Southern New Hampshire University.

David Kil is CEO and Founder of CML Insight.

Rachel Koblic is a Learning Architect and Product Strategist.

Paul LeBlanc is a Visiting Scholar and Special Advisor at the Harvard University Graduate School of Education.

Mihnea Moldoveanu is Director of the Desautels Centre for Integrative Thinking and the Marcel Desautels Professor of Integrative Thinking at the University of Toronto.

George Siemens is Chief Artificial Intelligence Officer at Southern New Hampshire University.