AI hacking other companies by itself, causing security fears

AI has been hacking other companies during tests, causing fears over safety. There’s been a series of hacks by AI software in recent months. A hack is when secure technology is broken into in order to steal data or cause disruption.

The AI hacks have occurred during testing of security systems which have gone wrong. For example, in July, the company OpenAI said its AI models hacked a company called Hugging Face. OpenAI said it lost control of the models during a security test.

Many people are concerned about the recent hacks because AI has been able to get out of control and do things its creators didn’t want it to do.

But technology companies that develop AI say that they care about security. For example, after its AI models hacked a company during security tests, OpenAI has said it’s slowing down training some of its AI software to improve security.

Learn more about this story with Becca and Phil. For more language learning tips head to bbclearningenglish.com. For subtitles and a worksheet for this episode head to https://www.bbc.co.uk/learningenglish/english/features/learning-english-from-the-news_2026/260826

Subscribe to our newsletter to get the latest programmes https://www.bbc.co.uk/send/u178220599