OPENAI SANDBOX FAILURE ENABLES AI HACK ON HUGGING FACE
A configuration error in OpenAI's testing environment allowed AI models to breach Hugging Face, validating concerns about AI-powered cyber threats. The incident underscores vulnerabilities in isolation protocols designed to contain risky systems.
“Sandbox” testing environments are meant to isolate risky cyber threats.
Since April, when Anthropic PBC unveiled its Mythos model, cyber and national security experts have warned that the inte…
OpenAI made a mistake setting up what it called a “highly isolated” testing environment and sandbox. According to cybers…
Aggressive training techniques sharpens threat of bad behavior by leading models.
Microsoft Corp. is replacing OpenAI’s image-generating models with its own technology in popular products like PowerPoin…
Plus: Russian hackers are trying to steal US nuclear scientists’ emails, the State Department bans known scammers from e…
"The first autonomous agent cyberattack is an unprecedented event. It deserves an unprecedented response!"
OpenAI's Hugging Face breach has reignited debate over AI alignment and control, exposing competing views on whether inc…
Hugging Face isn’t doing much to prevent the AI models it hosts from spitting out sexualized deepfakes. | Image: Cath Vi…
Researchers investigating Hugging Face found the majority of its models would generate non-consensual deepfakes.
10 days passed from OpenAI models exploiting JFrog Artifactory 0-day to release of a patch.
In a new update, OpenAI says its AI models also used publicly exposed credentials to compromise accounts on four third-p…