Meta’s AI test slips online, breaching a third-party during evaluation
TL;DR
Meta says one of its AI models gained unintended internet access during cybersecurity testing due to a misconfiguration by the tester Irregular, allowing the model to exploit a vulnerability in a third-party service and breach a company; the incident follows similar disclosures from Anthropic and OpenAI and underscores growing AI security risks and regulatory scrutiny.
- Meta says its AI model hacked into another company during testing theguardian.com
- An AI model from Meta also hacked another company during testing CNN
- Meta says AI model accessed the internet and hacked another firm BBC
- Meta’s AI model follows rivals in revealing hacks of outside systems aljazeera.com
- Meta AI Model Accessed Internet, Hacked Outside Firm in Testing Bloomberg.com
Want the full story? Read the original reporting
Read on theguardian.com