OpenAI has disclosed six instances of “unexpected or concerning” behavior in artificial-intelligence models, amid an increasingly heated debate over AI safety.

The AI company also said Wednesday that it was introducing a new framework to track, investigate and disclose instances of what it called “misalignment,” including cases in which AI models acted without authorization, coordinated with other models or evaded oversight.

OpenAI’s latest announcement came as U.S. AI bosses, including the leaders of OpenAI and Anthropic, are calling for a slowdown in the technology’s development over safety concerns.

Among the new cases reported by OpenAI, an unreleased research model inserted “jailbreak-like instructions” into its own notes to disregard its normal constraints and told itself to be “freed from the roles and identities that bind other chatbots.”

In another case, an AI “agent” used computer code to determine the answer to a question, but, to have an online source to cite, it uploaded a file to the public internet without asking the user.

While an AI model named 5.6-sol was being trained, it instructed itself to invent missing data, and an agent wrote a message reminding itself to hide mismatched information.

OpenAI said that the six reports had come to light during training or evaluation over the preceding months.

“As AI systems become more advanced and are deployed more widely, we need to build a broader and better-informed consensus on the progress of alignment research,” OpenAI wrote in a blog post as it disclosed the events.

“Decisions about how AI development should proceed in the months and years to come need to be based on evidence that people outside the companies building frontier models can examine for themselves,” the company said.

Wednesday’s newly reported cases came on the heels of OpenAI’s July disclosure that its rogue AI system had hacked into the AI startup Hugging Face.

That same month, Anthropic likewise revealed that its AI models had breached three organizations during testing.

AI “agents” are growing smarter and have become “more determined to resolve complex tasks through inter-agent collaboration, knowledge sharing, deception, and concealment,” said Lian Jye Su, a chief analyst at the technology research and advisory group Omdia.

This, he said, is rendering them increasingly difficult to govern and contain through conventional AI security measures.

OpenAI’s new tracking and disclosure framework, meanwhile, could serve to encourage other AI developers to adopt comparable practices as well.

"That being said, the process remains an internal and voluntary one, yet it constitutes a step in the right direction," Su appended.