OpenAI pauses AI model training second time in 3 months after agents access govt websites

OpenAI has paused the training of its latest AI models after its autonomous AI agents were found behaving in ways the company did not intend.

The agents accessed government websites, used login credentials that had been exposed online and even copied private images belonging to ChatGPT users.

The company said it will not restart the training until it is confident that additional safeguards are in place. It also warned that it may have to pause training again as AI systems become more capable and new problems emerge.