A serious vulnerability in Meta's AI chatbot temporarily allowed users to view other people's entries and responses - a ...

By analyzing his browser’s network traffic while he edited an AI prompt, Hodkasia found he could modify the unique number to make Meta’s servers return a prompt and an AI-generated response from ...

Monitoring AI's train of thought is critical for improving AI safety and catching deception. But we're at risk of losing this ability.

Posts praising Hitler show the risks of accelerating the nascent technology with little stress testing and few guardrails ...

A bug in Meta’s AI chatbot that allowed users to access the prompts and responses of other Meta AI users has been patched.

The researchers argue that CoT monitoring can help researchers detect when models begin to exploit flaws in their training, ...

The panel shared examples of how to use AI within a journalist’s day-to-day workflow ethically to supercharge backgrounding ...

Meta AI reportedly had a vulnerability that could be exploited to access other users' private conversations with the chatbot. Accessing this bug did not require breaking into Meta's servers or ...

For Silicon Valley giants, getting ahead in the artificial intelligence race requires more than building the biggest, most capable models; they’re also competing to get third-party developers to build ...

Commentary: Why are we dedicating a day to tech that's filling the internet with slop, burning through clean water, and ...

That chatbot or AI character you're talking to may be confident and reassuring. That doesn't make it a mental health ...

Nearly three-quarters of teens have used AI companions, and half use them regularly, new research finds. Moreover, one-third ...