Ford spent the past three years bringing back 350 veteran engineers after discovering that AI-driven assumptions in design and quality control left gaps the company still needed humans to fix. Charles Poon, vice president of vehicle hardware engineering, said Ford wrongly believed that feeding design requirements into AI would be enough to produce high-quality products. The issue, he said, was not that AI was useless, but that crucial engineering knowledge had never been documented in a way the systems could learn from. The case lines up with broader survey data cited in the report. An international study commissioned by Orgvue and conducted by Vitreous World, based on responses from 1,163 C-suite leaders and senior decision-makers, found that 39% had cut staff because of AI adoption. Of that group, 55% later said the decision was a mistake. Another 23% said layoffs were based on broad assumptions about AI capabilities rather than a detailed review of what affected employees actually did. The report also points to IBM and Commonwealth Bank of Australia as examples of companies that pulled work into AI systems, then ran into limits around judgment, edge cases or operating realities. Additional survey data from Careerminds, Robert Half and forecasts from Forrester suggest that a meaningful share of AI-linked layoffs may be reversed, with rehiring costs in some cases exceeding the savings from cuts.