Tech giants curb AI tool usage as costs surge

Large technology firms are restricting employee access to generative‑AI tools after confronting rapidly rising operational expenses. Meta announced limits on AI usage, Uber imposed monthly caps of roughly $1,500 per employee, and Walmart introduced token quotas for its internal AI agents. The shift follows a move from “tokenmaxxing” – encouraging maximal AI interaction – to “tokenmining,” which emphasizes cost‑effective deployment of cheaper models for routine tasks.

Analysts note that many companies suffered from “AI sprawl,” where individual teams adopted multiple chatbots, code assistants, and content generators, leading to duplicated effort, unpredictable costs, and auditing challenges. A Business Insider survey of 6,000 digital workers in the US, UK and Australia found that 77 % use several AI tools weekly, yet only 13 % see a clear boost in overall company performance.

Medium‑sized enterprises are also accelerating AI adoption, expecting cost reductions of 15‑30 % and productivity gains, while the finance sector increasingly relies on AI for portfolio construction and predictive analysis. Parallel developments include AI‑powered assistants for freelancers from Meta and Alibaba, expanding automation beyond large corporations.

These trends highlight a broader industry reassessment of AI’s value versus expense, prompting firms to prioritize impact‑driven use cases and tighter financial oversight.