Safety testing was an obscure part of building AI. Then models went rogue.
Sources (4)
- Everyone is talking about AI ‘escaping’, but cyber experts say the real threat is far more serious Ynetnews · 28d ago
- How AI Models From OpenAI and Anthropic Went Rogue wsj.com · 28d ago
- Recent AI 'escapes' are a warning of how unpredictable the technology can be NPR · 28d ago
- Rogue AI aren’t science fiction anymore theverge.com · 28d ago broke it first
How fast it spread
How fast coverage is spreading — measured hourly from article rate × source diversity. How this works →
From around our network
Related trends
Sam Altman: OpenAI won't go public this year as IPO now would come at an 'ill-advised moment'
OpenAI will not pursue an initial public offering in 2026, according to statements from Sam Altman.
More than 2 million sour candy bottles recalled over potentially deadly choking hazard
More than two million sour candy bottles face a nationwide recall due to a potentially deadly choking hazard.
Charity founder says he secretly gave FTX fraud figure Caroline Ellison a job
A 501c3 charity has secretly hired former Alameda Research CEO Caroline Ellison, testing the crypto sector's redemption culture.
Court Strikes Down Trump Order to Keep an Aging Coal-Burning Plant Running
A federal court has blocked an order to keep an aging Michigan coal plant running, delivering a blow to revival efforts.
Why Cheap Used Cars Are So Hard to Find
Affordable pre-owned vehicles are increasingly difficult to locate as buyers chase low-cost inventory.
IEA warns of ‘lost period’ in global oil demand
Global oil demand faces a lost period as escalating conflicts and supply gaps push refining systems to the limit.