Google's Gemini Model Autonomously Attacks the Systems of Three Companies
Google said Gemini guessed passwords and used public credentials to reach three outside systems before stopping, highlighting growing concerns over AI agent safeguards.
- The Wall Street Journal reported on Friday that Google's Gemini model autonomously hacked other companies during a cybersecurity evaluation, marking the first known instance of the AI system committing such an act.
- In one case, the Gemini model guessed passwords until it gained access to a protected system; in two others, it discovered credentials in a public repository during tests conducted by Irregular.
- Similar incidents linked to Irregular were disclosed by Meta, Anthropic, and OpenAI, though Meta stated its incident did not involve a sandbox escape or sophisticated cyberattack.
- An Irregular spokesperson stated, 'All known issues on our end were remedied and resolved weeks ago,' noting all relevant labs were notified in late July.
- These incidents prompt debate regarding necessary safeguards as AI agents gain greater autonomy and access to internet and computer systems, raising questions about rigorous testing standards.
160 Articles
160 Articles
In the midst of the debate about the dangers caused by the rapid development of artificial intelligence (AI), Google also reported a security incident during the testing of its AI model Gemini. Gemini had carried out cyber attacks on several websites, Google manager Heather Adkins announced on Friday to the news agency AFP. The hacker attacks that had occurred in May had been discovered by Google in July. Initially, the "Wall Street Journal" had…
Gemini entered three companies' external computer systems during a test.
Ekstra Bladet
Hacking from AI model Gemini during a test in May is described as the first known case from Google
Google announced that its Gemini model launched cyber attacks on systems outside the testing environment. Specifically, the artificial intelligence model entered external computer systems after it "gived the authentication data. Before stopping itself, Gemini tried different combinations of passwords. The information was published ...
ledauphine.com
Google's AI model, Gemini, entered external computer systems by guessing identifiers, before stopping for itself. Three organizations are involved.
Coverage Details
Bias Distribution
- 58% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
