Early evidence from companies and developers suggests artificial-intelligence tools are rapidly reshaping white-collar work. Venture firm a16z says chief technology officers now credit AI with a 30%–50% boost in output, and an internal team claims that 90% of its code is machine-generated. A separate online poll found that a majority of programmers already rely on AI for almost all of their coding. Corporate leaders are reacting by trimming head-count. IBM has replaced hundreds of human-resources employees with software as part of an 8,000-position reduction. Microsoft, which says as much as 30% of its code is written by AI systems, recently dismissed 9,000 workers. Buy-now-pay-later firm Klarna attributes a 40% workforce cut partly to automation, and Amazon chief executive Andy Jassy told staff he expects AI to shrink the retailer’s corporate ranks over time. The labour-market impact is starting to surface in macro data: Andrew Yang points to a July U.S. payroll gain of just 73,000—one of the weakest three-month stretches since the pandemic—and blames both tariffs and AI for the slowdown. Anthropic co-founder Dario Amodei forecasts that half of entry-level office jobs could disappear and unemployment could reach 20% within five years unless the economy adapts. Policy experts remain divided on how to respond. Economists David Autor and James Manyika argue in The Atlantic that regulators and industry should design AI systems that collaborate with, rather than replace, professionals, citing medical diagnostics as an example. Others, including University of Virginia professor Anton Korinek, warn that artificial general intelligence could arrive within five years, forcing a radical overhaul of social-safety nets. Japanese academics, meanwhile, have raised ethical flags about services that use AI to ‘recreate’ the deceased without consent. With productivity gains mounting and layoffs spreading beyond the technology sector, governments and companies face mounting pressure to balance the efficiency promised by AI with safeguards for jobs, skills training and human oversight.