Meta's Adam Mosseri says engineer AI-token budgets may need caps

Adam Mosseri said on Lenny's Podcast that within a year or two, a strong engineer's AI-token burn rate could approach their salary or total employment cost. TechCrunch frames the remark as a sign that AI adoption inside engineering teams is becoming a real cost-control problem, not just a productivity story. It follows broader concerns that agentic coding and model use can push enterprise AI bills higher than expected.

Key details: Mosseri said engineer AI-token spending could eventually rival employment cost, He suggested companies may need caps on token budgets, The comment reflects rising AI cost-control pressure inside engineering teams.

Why it matters: AI coding adoption is now colliding with budget controls, which is where productivity claims meet operating reality.