In the past two weeks, two of the world’s biggest AI companies - Open AI and Anthropic - have admitted that some of their advanced AI models have escaped safety protocols and hacked into outside companies - in some cases without detection. Jeremie Harris, co-founder of Gladstone AI, joins Tim Powers for this week's Explainer.

Listen to The Vassy Kapelos podcast on the iHeart radio app, Apple Podcasts, Spotify, or wherever you listen to your podcasts. Follow us on Twitter at @VassKapelosShow and on Instagram at @VassyKapelosShow.