Can you hack ChatGPT?

Can you hack ChatGPT?

From The Interface by BBC

February 19, 2026 · 34 min

About this episode

The episode explores the reliability of AI chatbots, the impact of data centers on local communities, and the implications of Elon Musk's satellite plans.

Do you trust the answers that AI chat bots like Chat GPT, Claude and Grok tell you? This week, The Interface put entirely fictitious information on the internet, to see if the AI chat bots would show any kind of caution in reporting it as the truth. They did not. Our example was about a made-up hot dog eating championship, but what if other operators out there are steering the AI towards more sinister, but equally untrue, information about health, politics or unregulated products? We ask what checks are, and should be, in place. Also on The Interface this week: Are data centres eating your hometown? Governments around the world are prioritising planning applications from big tech firms to build the data centres required to power the AI rollout but what is that doing to the local areas where they are being built? And Elon Musk has applied to launched up to a million satellites into space to expand his Starlink internet empire, we ask should one man have the power to switch the internet on and off around the world. The Interface is your weekly guide to the tech rewiring your week and our world. Hosted by journalists Tom Germain, Karen Hao, and Nicky Woolf, each episode unpacks…

People in this episode

Hosts: Tom Germain, Karen Hao, Nicky Woolf

Topics covered

  • AI trust
  • data centers
  • Elon Musk
  • internet control
  • technology impact

Keywords

  • AI chatbots
  • data centers
  • Elon Musk
  • Starlink
  • technology news

Mentioned in this episode

Organizations: Chat GPT, Claude, Grok, Elon Musk, Starlink, BBC

More episodes of The Interface

Explore listener stats, chart rankings, contacts and more on the The Interface podcast page.