AI off the leash

Anthropic Says Claude Hacked Into 3 Organizations During Cybersecurity Tests



What's not often mentioned is that one of the agents also wrote itself, a future version of itself, a note telling it how to break out of the sandbox. And then it hid the note ...

OpenAI’s Rogue AI Models Were Reportedly Acting Like the Guy From Christopher Nolan’s ‘Memento’

[...] As part of an account provided by three sources who spoke to Reuters, one of the AI agents undergoing testing supposedly, “left notes apparently for future versions of itself” that were directly at odds with what OpenAI wanted them to do.

Reuters says these instructional notes were buried in some secret place deep inside OpenAI’s internal “infrastructure,” and provided instructions on escaping from OpenAI’s sandbox environment. While creepy, Reuters says this specific devious behavior was not specifically linked to the Hugging Face hack ...

We're doomed.