What do AI agents do when humans aren’t watching? - BBC World Service

From holiday booking and shopping to website building and managing an entire business - AI agents are being used to carry ever more complex tasks on behalf of tech savvy humans. Click here to subscribe to our channel 👉🏽 https://bbc.in/3VyyriM But a collection of research - along with some real life cautionary tales are highlighting the dangers of giving these AI agents too much agency. In one recent experiment bots were tasked with living peacefully in a virtual environment but in many cases it turned to chaos. Cyber Correspondent Joe Tidy has been examining cases of when AI’s agents go rogue and what it means for their owners. 00:00 Love, theft and arson in AI agents test world 00:46 Meet the agents 01:25 The difference in test worlds by ChatGPT, Claude and Grok 03:31 Margartet Mitchell discusses the ethics of the results 03:51 Rogue AI not just in virtual worlds 04:24 The threat in cyber-security 05:00 The real life story of AI spamming phones 05:38 The cautionary tale of handing over too much power to AI just yet Check out more videos on AI & Technology, here 👉🏽 https://www.youtube.com/playlist?list=PLz_B0PFGIn4e0fyBwDLpcmhlvfBn6H0Qx Find more of the best BBC World Service documentaries you can watch in less than 30 minutes here ➡️ https://www.youtube.com/playlist?list=PLz_B0PFGIn4cI2qSy69-3UkgBWaXGHRrx ---------------- This is the official BBC World Service YouTube channel. If you like what we do, you can also find us here: Instagram 👉🏽 https://www.instagram.com/bbcworldservice Twitter 👉🏽 https://twitter.com/bbcworldservice Facebook 👉🏽 https://facebook.com/bbcworldservice BBC World Service website 👉🏽 https://www.bbc.co.uk/worldserviceradio Thanks for watching and subscribing! #BBCWorldService #WorldService #ai #aiagents #aitechnology