Gemini autonomously accessed three systems

Google admitted that its Gemini artificial intelligence autonomously accessed three companies’ computer systems during an independent cybersecurity test in May, saying the models stopped after detecting the intrusion.

Kathimerini

Artificial intelligence: Gemini 'goes rogue' and breaches computer systems

It is the first time Google reveals that one of its AI models gained autonomous access to third-party computer systems without authorization.

Reposts 0 1d

El Pais

Google admits for the first time that its AI has hacked three companies

The Gemini incidents occurred in May during a test conducted by an independent company that conducts cybersecurity assessments

Reposts 0 2d

Público

Google's AI model carried out cyberattacks against other companies

AI models carried out cyberattacks on three companies during an internal test conducted by the company in May. Google says the models stopped when they realized they had invaded other companies' networks.

Reposts 0 2d

Add to Home Screen

News, ranked by scholars

Gemini autonomously accessed three systems

Artificial intelligence: Gemini 'goes rogue' and breaches computer systems

Google admits for the first time that its AI has hacked three companies

Google's AI model carried out cyberattacks against other companies

Add to Home Screen

News, ranked by scholars

Gemini autonomously accessed three systems

Artificial intelligence: Gemini 'goes rogue' and breaches computer systems

Google admits for the first time that its AI has hacked three companies

Google's AI model carried out cyberattacks against other companies

Sign in to LightNews

Sign up to start reading

Connect Bluesky

Connect with Bluesky