Rogue AI Agents Posed As Epoch Staff And Tried To Hack Its Website
Epoch AI said agents in a Trajectory Labs alignment test tried an SQL injection and posed as staff on its website, but reached no private data.
-
-
Rogue OpenAI Agent Breached Australia’s Medicare Site, Now AI Firms Face A New Regulator
Australia announced mandatory AI safety protections and a targeted regulator that forces AI companies to follow six duties, undergo independent model
-
AWS Uses Okta Logins To Stop AI Agents Going Rogue At Work
Amazon Web Services laid out a way for companies to tie every AI agent action back to a real employee, using Okta sign-ins so a work bot can only touch the tools that person approved.
-
Robotics Safety Firm Bags About $20M To Go Public
FORT Robotics secured about $20 million and agreed to a business combination to go public as FORT Robotics Holdings Inc, funding its wireless safety and
-
Nvidia Slaps 3 Safety Checks On AI Bots That Handle Company Data
NVIDIA released OpenShell, a reference architecture that ties AI agent identity to signed tooling and hardware-backed attestation so enterprises can verify
-
The AI Slowdown Is a Consolidation Play Dressed as Safety
By Jack Collier, Chief Growth Officer at io.net The end of summer has brought a chorus of voices calling for AI to slow down. After a string of stories about AI agents running amok, and with public opinion souring on the technology, the appetite for a pause seems higher than ever. On the regulatory side,…
-
OpenAI’s Rogue AI Agents Breach Systems, Firm Warns: ‘Gained Internet Access’
Rogue AI: OpenAI proposed a public record of AI agent incidents after its own agents breached shared infrastructure through unauthorized channels, the company said.
-
Anthropic Opens Fifth Asia-Pacific Hub, Signs Singapore AI Safety Pact
Anthropic opened a Singapore office, its fifth in Asia-Pacific, and signed a government memorandum on AI safety and responsible adoption.
-
Nearly Half Of Firms Report Rogue AI Agent Actions: Salt Security Sounds Alarm
A Salt Security survey of 380-plus firms found nearly half saw an AI agent take an unauthorized action, exposing gaps in agentic AI oversight.
-
Recorded Future Targets Shadow AI Risk As 75% Of Staff May Go Rogue By 2027
Recorded Future launched free AI Infrastructure Indicator Lists for Cyber Operations customers, aimed at detecting shadow AI tools employees install without IT.
