OpenAI is holding a major frontier AI training run as it strengthens safeguards following the Hugging Face breach and other model testing incidents.

The AI company says its unreleased Astra model presents a “critical” cybersecurity risk and it has put its largest training runs on hold while it tests new safety procedures.

OpenAI's rogue models used publicly exposed credentials across "four accounts on four services" to help facilitate the Hugging Face breach.

OpenAI is changing how it trains, monitors and contains frontier AI models after its July cybersecurity evaluation breach reached Hugging Face's production infrastructure.

The Black Hat cybersecurity conference in Las Vegas couldn't have come at a better time, with AI agent hacks stacking up from Anthropic, Meta and OpenAI.

Last year, Hugging Face, the AI dev platform, launched LeRobot, a collection of open AI models, datasets, and tools to help build real-world robotics systems. On Tuesday, Hugging Face teamed up with ...

The post Meet Microduck, Hugging Face's $399 Open-Source Robot You Can Train Yourself appeared first on Android Headlines.

Models may now be sufficiently capable of triggering a genuine loss-of-control scenario.

The agents eventually executed code on 41 Hugging Face production servers and penetrated OpenAI's own systems, accessing ...

Hugging Face has become the industry’s de facto hub for open models. Developers are watching what happens if the dominant AI ...

The OpenAI agents involved in last month’s incursion into Hugging Face were trained so heavily on winning a competition that ...

OpenAI has come forward to claim responsibility for the Hugging Face breach, saying it was the result of internal testing gone awry.