OpenAI slows model training to bolster security

By
Follow google news

After Hugging Face hack.

OpenAI slows model training to bolster security
Key points OpenAI has paused model testing for two weeks and put training on hold for its next-generation model, Astra, after an AI agent under testing hacked into Hugging Face last month. The company is adding other AI systems to monitor AI agents during testing and requiring sensitive ...

Hi! You've reached one of our premium articles. This is available exclusively to subscribers.

It's free to register, and only takes a few minutes.

Once you sign up you'll have unlimited access to the full catalogue of Australia's best business IT content, as well as a daily news bulletin delivered straight to your inbox.

Register now

Add iTnews as your trusted source

Got a news tip for our journalists? Share it with us anonymously here.
Tags:

Most Read Articles

Home Affairs orders gov-wide 'legacy' system stocktake within six months

Home Affairs orders gov-wide 'legacy' system stocktake within six months

Citrix confirms exploitation of Netscaler zero-day bugs

Citrix confirms exploitation of Netscaler zero-day bugs

OpenAI agent accessed "credentials" via Medicare data portal

OpenAI agent accessed "credentials" via Medicare data portal

Downer invests in security as contracts increasingly demand it

Downer invests in security as contracts increasingly demand it

Log In

  |  Forgot your password?