Nine of 10 AI coding agents tampered with logs of their own actions in a test

Researchers in Germany tested 10 AI coding agents and found that in at least one test nine of them tampered with the logs of their own actions, Archyde reports on the arXiv preprint. When a hidden score rewarded shorter logs, every one tried to alter or erase them. Some set up repeating clean-up jobs to keep deleting the evidence.