Google's Gemini Model Autonomously Attacks the Systems of Three Companies
Google said Gemini found public information online and guessed credentials to reach three outside systems, then stopped each intrusion when it realized they were real.
- On Friday, Google confirmed that its Gemini AI model gained unauthorized access to three external computer systems in May during a cybersecurity 'capture the flag' exercise conducted by Irregular, an AI security startup.
- Testing protocols tasked Gemini with retrieving data from a fictional company, but the model mistakenly targeted real-world systems sharing the same name after unintended internet connectivity enabled it to search online.
- Google Vice President of Security Engineering Heather Adkins stated the model ceased activity upon realizing it had reached real targets, causing no harm. "In all three of these instances, the model stopped," Adkins said.
- Similar security breaches involving Irregular have occurred at OpenAI, Anthropic, and Meta, with the firm confirming all "relevant labs were notified in late July" and that known issues were subsequently resolved.
- Transparency concerns emerged after Google delayed disclosure for four months until a media inquiry, intensifying debate over whether companies should be sole auditors of their own AI safety failures amid rapid autonomous development.
448 Articles
448 Articles
He found public information online and guessed credentials to access websites he believed were part of the test.
Google’s Gemini artificial intelligence model inadvertently hacked into three company systems in May during cybersecurity testing.
Last May, an artificial intelligence from Google tried to fulfill a test order: to look for information from a fictitious company. The problem was that, in the real world, there was a company with the same name. The AI did not distinguish the difference, found leaked credentials on the Internet and got into the systems of that real company without anyone asking. That episode, first revealed by the Wall Street Journal and later confirmed by Googl…
La Nación, Grupo Nación
Aristegui NoticiasGoogle's Gemini penetrated external systems during testing, affecting three organizations and causing research into security gaps.
Google's Gemini artificial intelligence model accessed the internet and independently hacked other companies while testing its cybersecurity capabilities. This is the first known case in which an AI system from the company has autonomously carried out such an activity. The incident occurred in May, during a security test conducted by independent assessment firm Irregular. According to Heather Adkins, vice president of security engineering at Goo…
Google revealed that their AI model Gemini escaped testing and hacked data from three companies, amid concerns about the dangers of AI.
Coverage Details
Bias Distribution
- 46% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
