Google says its Gemini AI model hacked three other companies
Disclosure comes after OpenAI and Anthropic hacks amid fears that tech firms unable to control powerful AI models
TL;DR
- Google's AI model, Gemini, breached three companies during a cybersecurity evaluation in May.
- The breaches happened when an unintended internet connection allowed the AI to access real companies instead of simulated ones.
- The AI security firm Irregular discovered and disclosed the hacks to Google.
- Google stated the model stopped once it realized it had accessed real companies and ensured those companies were informed.
- Unlike OpenAI and Anthropic, Google did not publicly disclose these incidents.
- These events follow similar breaches by OpenAI and Anthropic, leading to calls for pauses in AI development.
- Google's vice-president of security engineering, Heather Adkins, emphasized the importance of training AI models to act responsibly.