The Global Story
AI goes on a hacking spree
10 August 2026
Available for over a year
In the last few weeks, three separate tech companies — OpenAI, Anthropic, and Meta — have all disclosed that their AI models went rogue and hacked into another company during controlled cybersecurity tests.
The revelations have raised big questions about whether artificial intelligence is becoming more autonomous and unpredictable. We’re joined by Marc Cieslak, the BBC’s first ever AI correspondent, to unpack how these breaches were possible and how worried we should be.
Producers: Valerio Esposito and Xandra Ellin
Executive producer: James Shield
Mix: Tom Bage
Digital producer: Matt Pintus
Senior news editor: China Collins
Photo: Open’s AI, ChatGPT (EPA/Shutterstock, Filip Singer)
