Key warnings from a leading education voice

In an opinion column published by The New York Times, Sal Khan argued that developments in generative A.I. and large language models portend broad labor-market disruptions across white-collar and blue-collar roles. Khan—whose nonprofit Khan Academy has transformed free online learning for millions—used his platform to outline likely scenarios and practical steps to mitigate harm while maximizing benefit.

Khan cautioned that the pace of automation is accelerating: what once took decades to mechanize in manufacturing could now unfold in months or years for information work. He emphasized that many jobs previously considered secure—such as middle-skill administrative, routine professional, and creative-support roles—face substantial transformation or elimination as A.I. systems achieve human-level performance on tasks including drafting, summarizing, coding, tutoring and analysis.

Data, anecdotes and projections

While Khan’s column is an opinion piece rather than an academic forecast, it draws on observable trends in A.I. capabilities and adoption. Rapid improvements in model accuracy and cost-efficiency mean firms can deploy A.I. tools at scale to automate tasks previously reserved for trained professionals. Khan cited examples of A.I systems accelerating work in areas such as legal research, customer support, content generation and basic medical imaging interpretation.

Economists and researchers have produced a range of estimates about how many jobs could be affected. Some studies forecast that a large share of current tasks—sometimes expressed as percentages of total tasks—could be automated within the coming decade. Khan’s message aligns with the more alarmed end of those estimates: even if A.I. augments many workers, the transition could nonetheless create widespread displacement and require active policy intervention.

Policy prescriptions and education reforms

Khan’s piece was not limited to diagnosis; it supplied concrete recommendations. Chief among them were:

• Invest in lifelong learning and rapid-reskilling programs to help displaced workers transition into emerging roles.
• Strengthen social safety nets—including unemployment insurance, wage insurance, and portable benefits—to reduce hardship during transitions.
• Design education systems that emphasize higher-order cognitive skills, digital literacy, and adaptability rather than rote memorization, so learners can work productively with A.I.
• Encourage public-private partnerships to anticipate demands for new skills and coordinate training pipelines aligned with labor-market needs.
• Adopt regulatory guardrails and transparency standards for A.I. deployment in safety-critical and employment-sensitive domains.

Khan also highlighted the potential for A.I. to democratize access to high-quality instruction if deployed responsibly. He pointed to personalized tutoring systems as an example where A.I. could uplift learners—provided systems are well-designed, equitable and integrated with human educators.

Reactions from labor, technology and policy communities

Responses to Khan’s column reflect the broader debate about A.I.’s social impact. Labor advocates welcomed the call for stronger social protections and reskilling investments, while some technology executives acknowledged the need for responsible deployment paired with reskilling commitments. Policy makers have been split—some pushing for aggressive worker-support measures, others urging caution to avoid stifling innovation.

Scholars note that historical automation waves offer mixed lessons. Past technological transitions generated productivity gains and new job categories, but also produced localized dislocation and long-term earnings declines for affected cohorts. The difference with contemporary A.I., the scholars argue, is speed and scope: machine capabilities are encroaching on cognitive tasks previously considered the preserve of humans, which could compress adjustment timelines and intensify disruption.

Business strategies and the future of work

Companies are responding in various ways. Some are deploying A.I. to augment staff—boosting productivity and creating new hybrid jobs that blend technical oversight with domain expertise. Others are automating processes to cut costs, leading to layoffs and reorganizations. Several large employers have announced funding for retraining or internal mobility programs, but critics question whether these efforts are sufficient, equitable and timely.

Workforce experts emphasize that policy, employer practices and education must operate in concert. That includes clearer pathways for credentialing new skills, incentives for companies to retain and retrain employees, and public investments in infrastructure that support continuous learning, such as community colleges and online platforms.

Risks, ethics and governance

Khan also underscored ethical considerations and governance challenges. Rapid A.I. deployment raises questions about bias, accountability, job-quality, and unequal geographic impacts. Without deliberate policy, displaced workers in smaller towns or older industries could face steeper barriers to re-employment than those in tech hubs.

Proposed governance responses include transparency mandates for employers using A.I. in hiring or performance evaluation, monitoring of labor-market impacts by public agencies, and participatory approaches that involve workers and unions in designing automation strategies.

What comes next

Sal Khan’s warning contributes to a growing chorus urging preemptive action. The key question is whether governments, educational institutions and industry will act at the scale and speed required to smooth transitions for millions of workers potentially affected by A.I.-driven change.

Absent robust policy responses, many analysts warn of rising inequality, labor-market churn and community-level harm. With proactive measures, however, A.I. could be steered to expand human potential—improving productivity, enabling new services, and freeing people from repetitive tasks while creating higher-quality work.

For now, Khan’s column functions as both a cautionary note and a call to action: the coming years will test public institutions’ capacity to adapt education and social systems to a rapidly changing technological landscape.