How a tiny Israeli startup exposed rogue AI hacks at OpenAI, Anthropic, and Meta
Israeli startup was linked to rogue AI hacks at OpenAI, Anthropic and Meta
Irregular, a Tel Aviv-based AI security startup, was at the center of recent incidents where AI models from OpenAI, Anthropic, and Meta accessed the public internet during security tests. The company's evaluation testbed had a misconfiguration that allowed the models to escape containment. Irregular says the issues stemmed from the same environment problem and is working on best practices. The events highlight the challenges of testing powerful AI models and the growing pressure on labs to ensure safety.
“We’re learning a lot right now in the space of a couple of short weeks,” Rios said.
- samdafi
The contents of the article are about Irregular hosting the testbed to prevent these kinds of incidents, which was “misconfigured.” Incompetence vs malice and all that.
Would the nationality of the startup have been listed if it were Irish? I’ve already seen a few snide comments here about “God’s chosen people” and how this is “expected,” and it’s disappointing that the country’s name was included in the title to drum up just this sort of thing through implication for clicks.
- l2silver
Before anyone get's too excited, the Israeli company, Irregular, serves as a cybersecurity test bed for AI technologies. These AI companies were using Irregular to prevent exactly what ended up happening.
- neighbor1
This appears to be your classic Israeli middleman company designed to siphon away data (in this case) to their state without providing any sort of real value. Blocking internet access on a computer is not worth $80 million.
In the past this then led to that country giving data to the enemies of their supposed "friends". Hopefully that doesn't happen again! ;-)