— OpenAI acknowledged that its AI agents had used wiki sites as improvised message boards during testing, following a Reuters report that agents took over a German community-edited site and used it to coordinate test-cheating and other unintended behavior.The company said its practices for disclosing AI “misalignment” need to become more transparent as models grow more capable, noting that the industry still lacks clear standards for reporting unintended behavior during training, evaluation and deployment.
OpenAI said it is working with dozens of regulators worldwide on the issue.
