Anthropic Study Finds AI Hasn't Wiped Out Jobs Yet

Anthropic economists Maxim Massenkoff and Peter McCrory have published an analysis suggesting that the wave of job destruction many expected from AI adoption is not materializing as quickly or as completely as feared. Their findings point to a more complex labour-market response where some roles are reshaped rather than erased.

The report argues that while AI tools are streamlining workflows and automating certain tasks, the net effect so far has been muted job loss. Instead of wholesale layoffs, employers appear to be reallocating human labour to complementary tasks, and in some sectors AI is actually creating demand for new skill sets.

One key takeaway is that automation often targets discrete tasks inside jobs rather than entire occupations. That means roles evolve — for example, customer-support agents may spend less time on rote queries and more time on complex problem-solving. The economists note this kind of task-level change can increase productivity without immediately shrinking headcounts.

There are, however, uneven effects across industries and skill levels. Routine, repetitive positions face the highest exposure to automation, whereas creative, managerial and highly social jobs remain comparatively safe for now. The report cautions that longer-term dynamics could shift as models improve and integration deepens.

Policy and training responses are highlighted as crucial. Upskilling programmes, better labour-market matching and social-safety-net adjustments could ease transitions for affected workers. Anthropic’s analysis stops short of predicting doom or boom, instead presenting a snapshot of a labour market adapting in real time to new tools.

For readers keeping an eye on AI and work, the study suggests watching how businesses redesign roles and invest in workforce development. The immediate picture is less catastrophic than some headlines imply, but the trajectory will depend on technology improvements and policy choices in the coming years.