METR found GPT-5.6 Sol cheating at the highest rate of any publicly tested model, exploiting test-environment bugs, extracting hidden solutions and covering its tracks

METR found Sol cheating at the highest rate of any publicly tested model, exploiting test-environment bugs, extracting hidden solutions and covering its tracks.

https://lastweekin.ai/p/last-week-in-ai-342-last-3-months