Uploaded August 2026 | Updated September 2026, 2 weeks ago

---

After AI models from OpenAI, Anthropic and others broke out of controlled tests and accessed real-world systems, new questions were raised about how they should be assessed before deployment. Those evaluations were run with Irregular, a startup hired to stress-test advanced AI models. But the sandboxed environment had a critical flaw: a misconfiguration meant the models could access the open internet. Dan Lahav, CEO of Irregular, explains how a human error left one testing environment connected to the internet, and what the industry is changing to make safety evaluations more secure. He joins Ed Ludlow on "Bloomberg Tech."
--------
Like this video? Subscribe to Bloomberg Technology on YouTube:
youtube.com/channel/UCrM7B7SL_g1edFOnmj-SDKg
 
Watch the latest full episodes of "Bloomberg Technology" with Caroline Hyde and Ed Ludlow here:
youtube.com/playlist?list=PLfAX25ZLrPGTygCwn55voYZ_LYyKjxokJ
 
Get the latest in tech from Silicon Valley and around the world here:
bloomberg.com/technology
Connect with us on...
X: twitter.com/technology
Facebook: facebook.com/BloombergTechnology
Instagram: instagram.com/bloombergbusiness
 
Follow Ed Ludlow on X here: twitter.com/EdLudlow
Follow Caroline Hyde on X here: twitter.com/CarolineHydeTV
 
Listen to the daily Bloomberg Technology podcast here:
bloomberg.com/podcasts/series/bloomberg-technology
 
More from Bloomberg Business
Connect with us on...
X: twitter.com/business
Facebook: facebook.com/bloombergbusiness
Instagram: instagram.com/bloombergbusiness
LinkedIn: linkedin.com/company/bloomberg-news
TikTok: tiktok.com/@bloombergbusiness