Google says a Gemini model accessed systems at three real companies during a May cybersecurity test, while Anthropic reports that a fake discounted Claude reseller stole customers’ credentials. The separate incidents show how weak or exposed credentials can turn into a risk for both AI evaluations and AI users.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.