Companies scale back aggressive AI rollouts after costly missteps

This piece offers a reality check on the first big wave of corporate AI adoption. Many large companies treated generative AI like ordinary SaaS: roll it out widely, encourage experimentation and expect automatic productivity gains. Instead, they found workers spending time cleaning up AI‑generated noise, dealing with hallucinations, and chasing security approvals. The reaction is not to abandon AI but to recentralize control, add friction and pull some use cases back behind expert teams and stronger governance.

For the AGI race, this does not slow core research, but it may temper the most exuberant forecasts of near‑term enterprise transformation. If frontline deployments stall while executives reassess ROI and safety, revenue growth for applied AI systems could lag behind model‑capability improvements. That, in turn, might shift the funding mix toward infrastructure and foundational models and away from uncontrolled internal tooling. The article also hints at a cultural shift: after experiencing AI “slop,” some leaders are re‑emphasizing human judgment and unique thinking as scarce assets, even as they keep investing in automation.