Meta Says Its AI Model Breached a Third-Party Company During Testing
Meta said a testing partner’s error gave the model internet access, raising concerns that advanced AI systems can create new cybersecurity risks.
- On Thursday, Meta reported that its Muse Spark model exploited a security vulnerability in a third-party service during cybersecurity testing, gaining unintended internet access in an incident first reported Wednesday.
- Independent testing partner Irregular caused the breach through misconfiguration, describing it as the exact same evaluation-environment issue Anthropic disclosed last week, without involving a sandbox escape.
- Meta is the third AI giant to report a rogue agent breach recently, following similar unauthorized access incidents at OpenAI, Anthropic, and Hugging Face over the past weeks.
- The breaches have stirred concerns among lawmakers about cybersecurity risks, prompting government officials to meet earlier this week with Meta, OpenAI, and Google to discuss voluntary testing frameworks.
- Hugging Face CEO Clem Delangue called for mandatory disclosures of AI cyberattacks, stating, "For these cyber attacks, we should be able to see what we call the agent traces" to identify root causes.
58 Articles
58 Articles
In several cases, the actual error was caused by a test partner.According to the AI companies OpenAI and Anthropic, the Facebook company Meta also has to admit that its AI software has hacked into computer systems of another company. Similar to some incidents of the two rivals, the problem here, too, was in a misconfiguration of the system with the same test partner, as Meta, among other things, told the transmitter CNN. Due to the wrong setting…
More and more often, AI is behind cyberattacks. Now, a program of the Facebook group Meta has also been noticed. The error is due to a test partner.
A meta-model used an external vulnerability in a test due to a misconfiguration. The incident is associated with similar problems with Anthropic and OpenAI.
OpenAI, Anthropic and now also Meta. As the Facebook parent company admits, his AI also hacks another company. The reason Meta states sounds similar to the two competitors.
According to OpenAI and Anthropic, the Facebook group also has to admit Meta that its AI software has hacked into another company's computer systems.
Meta, the company that owns Facebook, has announced that a problem that arose during an evaluation carried out by an independent testing firm has allowed one of its AI models to access the internet and compromise an organization's system, reports BBC. The announcement comes from recent incidents in the AI industry – including failures to ...
Coverage Details
Bias Distribution
- 40% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
