Union Budget 2026 | Education Sector Analysis
February 5, 2026AI Is Reshaping Jobs.
Here’s How Indian Education Can Lead the Shift.
On 5 March 2026, Anthropic published a study that does something most AI-and-jobs research hasn’t: it measures what people are actually using AI for at work, not just what AI could theoretically do. The findings carry a pointed message for Indian educators.
The Core Finding
The study introduces a metric called observed exposure — a measure that combines the theoretical capability of large language models with real-world professional usage data from millions of Claude conversations. The result is a far more grounded picture of which jobs are being touched by AI today, and which are not.
The headline conclusion: AI’s actual labour-market impact is still early. Overall unemployment in AI-exposed occupations has not risen. But beneath that reassuring surface, the data reveals a pattern that should concern anyone in education.
The Young-Worker Signal
Perhaps the most consequential finding for educators: hiring of 22–25-year-olds has slowed in AI-exposed occupations, even as overall unemployment in those roles remains flat. Young people are not being fired — they are simply not being hired at the same rate. Companies are discovering that AI can absorb work that used to be given to entry-level staff.
This is a leading indicator, not a lagging one. And it lands directly at the doorstep of colleges and universities whose primary promise is employability.
Why This Matters for India
The “safe career” list is shifting
Computer programming, financial analysis, customer service, data entry, and administrative roles sit at the top of Anthropic’s exposure index. These are exactly the career tracks that Indian families have invested in for decades. The entry-level version of these jobs — the version a fresh graduate performs — is the portion most likely to be absorbed by AI.
India’s entry-level economy is exposed
India produces over 1.5 million engineering graduates annually. Hundreds of thousands more graduate in commerce and management. Many enter the workforce through precisely the tasks the study identifies as high-exposure: basic coding, report writing, data entry, customer support. The IT services and BPO sectors, which employ millions in task-level work, sit squarely in the risk zone.
The gap is the opportunity. Only about a third of theoretically automatable tasks are currently being handled by AI. This gap gives Indian institutions a window — perhaps three to five years — to retool before the impact becomes unmistakable.
AI fluency is not optional
The report distinguishes between automated use (AI working independently) and augmented use (AI assisting humans). Most current usage is augmentative. The immediate demand is not for people who build AI, but for people who can work effectively with AI — in law, finance, design, healthcare, and research. Every college department, not just computer science, should be asking whether its graduates are learning this.
What Should Educators Do?
For schools: Move career counselling from fixed “prestigious course” ladders toward honest conversations about which skills within a field are durable and which are automatable. Introduce AI literacy as a cross-cutting theme, not a standalone elective.
For colleges: Audit curricula against the exposure data. Programmes heavy on procedural execution need the most urgent revision. Make AI-augmented coursework the norm — students should learn to use, evaluate, and improve AI output, not just produce work AI can already do.
For policymakers: Commission an Indian observed-exposure metric using domestic employment and AI-usage data. The Anthropic study uses US occupational categories; India’s workforce structure is different and needs its own measurement.
The Bottom Line
Anthropic’s study is careful to say the sky is not falling — yet. But the slowdown in young-worker hiring is a signal that arrives before the storm, not after it. For a country where education is the primary vehicle for economic mobility, waiting for the impact to become obvious is not a strategy.
The time to act is while the gap between AI’s capability and its adoption is still wide enough to prepare.
Read the Full Insight Paper
We’ve published a detailed insight paper with occupation-level data, specific recommendations for Indian schools, colleges, and policymakers.
Download the PDF →Source: Massenkoff, M. & McCrory, P. (2026). Labor Market Impacts of AI: A New Measure and Early Evidence. Anthropic.
https://www.anthropic.com/research/labor-market-impacts
