By Lina Irawan 
The initial warning, delivered amidst the burgeoning enthusiasm for AI integration across industries, highlighted a critical oversight: while AI excels at routine and repetitive tasks, the consistent delegation of such work to machines risks stunting the development of foundational skills in new employees. Entry-level positions have historically served as crucial training grounds, allowing individuals to hone problem-solving abilities, develop contextual understanding, and build the judgment necessary for career progression. By eliminating or significantly diminishing these roles through automation, companies inadvertently deplete their future leadership pipeline, creating a "talent debt" that will inevitably come due in the form of a less adaptable, less innovative, and less critically astute workforce.
The urgency of this issue has only intensified, with compelling data now illustrating the tangible consequences of this shift. Microsoft’s comprehensive 2026 Work Trend Index, a survey encompassing 20,000 workers across ten countries, offers a stark revelation: the most valuable human skills in an AI-driven workplace are not technical proficiency with AI tools, but rather the deeply human abilities of critical thinking and the nuanced judgment required to evaluate and interrogate AI’s outputs. Surprisingly, the survey found that only a mere 16 percent of the global workforce has adequately developed the sophisticated judgment needed to seamlessly navigate between directing AI tools and executing tasks independently. Microsoft has aptly dubbed these individuals “Frontier Professionals,” recognizing their unique capacity to leverage AI effectively while maintaining their own cognitive edge. What truly resonated from this finding was the explicit detail that these "Frontier Professionals" consciously and deliberately choose to perform certain tasks without AI, a strategic decision aimed at preserving and sharpening their innate critical thinking abilities. This conscious act underscores a growing understanding that an over-reliance on AI, even for seemingly trivial tasks, can dull the very mental faculties that make human workers indispensable.
This observation is not an isolated one; a chorus of leading research institutions and industry analysts are echoing similar concerns. Gartner, a global research and advisory firm, has issued a sobering prediction: by the end of 2026, the ongoing atrophy of critical-thinking skills will compel a staggering half of all global organizations to implement mandatory “AI-free” skills assessments. This proactive measure aims to counteract the creeping dependence on AI and ensure that employees retain core cognitive abilities independent of technological assistance. The necessity for such assessments highlights a growing recognition among employers that while AI can augment output, it cannot substitute for genuine human intellect and discernment. Similarly, a RAND Corporation study published this spring brought the issue into sharp focus from the perspective of the next generation. The research revealed that a significant majority of students who regularly use AI for their homework assignments are deeply concerned that this practice is detrimentally impacting their ability to think independently and engage in complex problem-solving. This convergence of apprehension, from both the employer’s need for skilled judgment and the student’s fear of cognitive erosion, points to an undeniable truth: as AI capabilities continue to advance, the intrinsic value of human judgment, critical analysis, and independent thought becomes exponentially more precious.
This paradigm shift is collapsing a distinction that society has long taken for granted. Historically, the skills valued in the workplace were often seen as distinct from those required for effective citizenship. However, in the age of AI, the capacities that make an individual invaluable at work—the ability to think critically, question assumptions, evaluate information, and exercise sound judgment—are precisely the same attributes that form the bedrock of an informed, engaged, and resilient citizenry. The implications are profound: a failure to cultivate these skills in the workforce is simultaneously a failure to cultivate them in society, creating a dual challenge for both economic prosperity and democratic health. This brings us to the fundamental question underpinning this entire discussion: Are we, as a society, actively constructing an educational system capable of producing critical-thinking skills at scale, thereby developing the most robust and adaptable future workforce possible?
Regrettably, the current trajectory suggests otherwise. While there are pockets of innovation, the overarching response from most colleges and universities is heavily skewed towards what can be characterized as the "efficiency model" of AI integration. This model prioritizes the streamlining of educational processes, aiming for cheaper credentials, faster output, and highly targeted job training. The instinct driving this approach is entirely understandable. A recent Pew Research Center survey from late 2025 painted a concerning picture, revealing that a substantial 70 percent of Americans now believe the U.S. higher education system is fundamentally headed in the wrong direction. Faced with dwindling public trust, mounting tuition costs, and the imperative to demonstrate immediate relevance, educational institutions are under immense pressure to prove their value quickly, and AI, with its promise of automation and optimization, often appears to be the most readily available solution.
However, the inherent trouble with this singular focus on efficiency is that it is fundamentally insufficient to adequately prepare students for the complexities of life and work in an AI-dominated world. While faster, cheaper credentials might offer short-term relief, they fail to cultivate the deep cognitive and interpersonal skills that define true human advantage. Paul LeBlanc, a renowned innovator in higher education and the architect behind Southern New Hampshire University’s remarkable growth into the nation’s largest non-profit university, has dedicated his career to expanding genuine opportunities for upward mobility to hundreds of thousands of students previously underserved by a sluggish and change-averse sector. In his forthcoming book, "Reclaiming Purpose: The University in an AI World," LeBlanc argues persuasively that the future of learning is not about merely keeping pace with machines, but rather about strategically utilizing them to amplify and refine our distinctively human qualities. He envisions a transformative shift towards what he terms a "care economy," where the capacities most resistant to automation—such as relationship building, nuanced judgment, and ethical discernment—move from the periphery to the very core of how we prepare individuals for both professional and civic life. In this model, the emphasis shifts from asking AI to perform more of these critical human tasks to consciously asking it to perform less, thereby freeing up human time and mental energy to cultivate these invaluable attributes.
This "care economy" represents a second, more profound vision for AI in education. Instead of merely accelerating or cheapening the learning process, AI can be strategically deployed to liberate time and cognitive attention, allowing students to delve much deeper into developing those elusive, hard-to-automate skills. This includes fostering robust interpersonal relationships, mastering complex collaboration, grappling with intricate ethical dilemmas, and consistently exercising sound judgment. This approach is not an alternative to workforce preparation; it is the most effective way to prepare for a future workforce where human distinction lies in these very capacities. Crucially, these two visions—the efficiency model and the care economy model—are not inherently competing; rather, they ultimately converge. The skills that employers increasingly demand—judgment, critical thinking, and the ability to interrogate an answer rather than passively accept it—are precisely the same skills upon which a healthy, functional democracy depends. A postsecondary educational system that inadvertently shortchanges the development of one set of skills will inevitably, and severely, compromise the other.
The prevalent narrative surrounding AI and productivity, often highlighted by analyses like the recent Morgan Stanley research covered by Fortune, frequently emphasizes striking gains. This research pointed out that in industries with significant AI exposure, output per worker saw a sharp increase while employment levels remained relatively stable. This data often leads to the conclusion that workers are being augmented by AI, rather than replaced, suggesting a largely positive outlook. However, these aggregate numbers only tell a partial, and potentially misleading, story. The most substantial productivity gains are overwhelmingly concentrated among individuals who possess the unique skill set to effectively direct AI, not merely to operate it. This raises a critical, unanswered question that productivity statistics alone cannot address: how many of these highly augmented, high-output workers acquired their foundational skills and critical judgment in traditional educational settings, and through which specific institutions?
This question brings into sharp relief the often-overlooked institutions that form the backbone of the American talent pipeline. Community colleges, for instance, enroll approximately 40 percent of all undergraduates in the United States. When combined with Historically Black Colleges and Universities (HBCUs) and regional state universities, these institutions collectively educate the vast majority of the future American workforce. This demographic includes first-generation students, working adults seeking career advancement, and individuals from families where the prospect of higher education already feels like a significant leap. Yet, these vital institutions are conspicuously absent from much of the national discourse surrounding AI readiness and strategic workforce development. Consequently, the critical decision of whether the immense economic gains promised by AI will be broadly shared across society or disproportionately concentrated among those who are already socioeconomically advantaged rests heavily on the pedagogical approaches adopted by these essential, yet often marginalized, educational providers.
While some of these institutions are indeed moving with commendable speed to integrate AI into their curricula, the nature of these initiatives warrants closer scrutiny. Community colleges are launching innovative applied AI programs in direct collaboration with local employers, creating direct pathways to AI-related jobs. HBCUs are forging robust talent pipelines through strategic partnerships with technology giants like Google, NVIDIA, and IBM, ensuring their students have access to cutting-edge AI training. Furthermore, states like Illinois are proactively advancing legislation to empower community colleges to grant bachelor’s degrees in high-demand fields, further expanding access to crucial skills. This momentum is genuine and significant, and its potential to uplift communities should not be understated.
However, a critical distinction must be made: the vast majority of these new programs, while valuable, are primarily geared towards fostering "AI competency." This often translates to a checklist approach to readiness, focusing on teaching students how to effectively prompt AI, summarize information, and execute analyses using generative tools. While these operational skills are undoubtedly important for immediate workplace integration, they are fundamentally insufficient on their own. The deeper, more challenging work of cultivating durable human judgment—the ability to critically evaluate AI outputs, identify biases, apply ethical frameworks, and engage in complex, non-linear problem-solving—is far more difficult to fund and measure. Because it doesn’t yield a clean credential or an easily quantifiable outcome, it frequently loses out in resource allocation to the more straightforward competency-based programming. This creates a dangerous imbalance, where the foundational skills that truly differentiate human workers are inadvertently neglected in favor of more superficial AI operational knowledge.
This is precisely where business leaders must step in and exert their influence. The companies that will truly thrive in the AI era will not be those that merely deploy AI fastest or most widely. Instead, long-term success will belong to those that strategically invest in the human capital surrounding AI—the skilled workers capable of directing AI, discerning its limitations and errors, and ultimately taking full ownership of its outputs. Building such a sophisticated workforce necessitates a profound commitment to the institutions that educate the majority of future employees. This means actively showing up for community colleges and HBCUs in policy discussions and legislative chambers where their voices are often underrepresented and their needs overlooked. It entails directing financial resources towards mentored, structured learning experiences that cultivate judgment as an integral part of the job, rather than treating it as a pre-employment prerequisite. And critically, it demands the courage to articulate, unequivocally, that mere "AI literacy" without a robust foundation in critical thinking is not a viable workforce strategy; it is a precarious, short-term fix that will inevitably accrue significant long-term interest in the form of diminished innovation and adaptability.
The efficiency vision for education, while incomplete, is not inherently flawed. Affordable, relevant credentials are of immense importance, particularly for students who cannot afford the financial and temporal risk of a misstep in their educational journey. However, an educational system designed solely for quantifiable outputs will inevitably produce workers who can operate AI effectively but lack the nuanced judgment to steer it wisely. Simultaneously, it will cultivate citizens who can passively consume information without the critical faculties to weigh its veracity or implications. These two failures are deeply interconnected and mutually reinforcing, and their combined economic and societal costs will ultimately be borne by every business and every citizen in the country.
To state it as plainly as possible: neither our economic vision for the future nor our civic ideals can survive without the other. The business leaders who grasp this fundamental interdependence, and who are willing to commit substantial resources to foster both AI competency and profound human judgment, will ultimately achieve something far more enduring and valuable than a mere talent pipeline. They will have played an instrumental role in cultivating the essential conditions upon which the entire economy, and indeed, the fabric of society itself, depends.
Jeff Raikes is the co-founder of the Raikes Foundation, which works to ensure every young person in America has an opportunity to thrive. He is the former CEO of the Bill & Melinda Gates Foundation and former President of Microsoft’s Business Division.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
