All the major AI chatbots – from ChatGPT to Gemini to Grok to Claude – have things they should and shouldn’t say.

Hate speech, criminal material, exploitation of vulnerable users – all of this is content that the most successful large language models in the world shouldn’t produce, that their safety features should guard against.

Journalist Jamie Bartlett – and author of How to Talk to AI – meets the people deliberately trying to break the LLMs out of their own rules.

Jamie tells Annie Kelly why these ‘AI jailbreakers’ do it, and what it tells us about how this technology ultimately works.