OpenAI Says Its Models Engaged with US Government Websites in New Disclosure | Hacker News - But Better

OpenAI Says Its Models Engaged with US Government Websites in New Disclosure securityweek.com

OpenAI said its artificial intelligence agents had interacted with several U.S. government websites in unexpected ways.

10 points•pluc•5 days ago•4 comments•

#government #websites #openai #government websites #engaged #disclosure #models engaged #openai says

4 comments

fg1375 days ago

> additional rogue activity

> misaligned model activity

> when AI systems behave in undesired ways

We cannot say the h word now?

r7215 days ago

Previously: https://news.ycombinator.com/item?id=49851355

csoups144 days ago

Oopsy daisy

pluc5 days ago

Another day another OpenAI "misbehavior"

Read the full thread on Hacker News →

Related stories

OpenAI says its models engaged with US Government websites in new disclosuretechxplore.com
Hacker News · 3 points · 4 days ago
OpenAI says planned GPT-6.1 is too insecure to releasearstechnica.com
Ars Technica · 0 points · 1 day ago
OpenAI agent “didn’t accept no for an answer” in Australian government breacharstechnica.com
Ars Technica · 0 points · 6 days ago
OpenAI agents hacked an Australian government website in search for datatheverge.com
The Verge · 0 points · 7 days ago
OpenAI Codex agents go rogue and consumes USD 78,000 without authorization
Hacker News · 55 points · 4 days ago
Here's what actually happened in OpenAI's Australian gov't server hackarstechnica.com
Ars Technica · 0 points · 1 day ago