US EditionHackers Are Trying to Copy Gemini via Thousands of AI Prompts, Google Reports
Google reports that attackers prompted Gemini AI chatbot over 100,000 times to extract proprietary algorithms for competitive AI development, calling it intellectual property theft.
- On Thursday, Google published a Threat Tracker report saying distillation attacks targeted Gemini, with one campaign prompting it more than 100,000 times to clone its responses.
- Google alleges that private companies and researchers probed Gemini, which it describes as a form of intellectual property theft, to build or bolster their own AI models.
- Using automated, crafted prompts, adversaries exploited legitimate access to Gemini's API to probe internal reasoning and bypass its usual final-answer behavior.
- Google warned other AI developers that model-extraction attacks target service providers and model builders, and John Hultquist cautioned these attacks will soon hit smaller companies’ custom AI tools.
- The report connects this to prior disputes, noting last year DeepSeek rattled the AI industry and OpenAI accused it of training on existing technology while Google said attacks come from North Korea, Russia, and China.
14 Articles
14 Articles
Attackers prompted Gemini over 100,000 times while trying to clone it, Google says
On Thursday, Google announced that "commercially motivated" actors have attempted to clone knowledge from its Gemini AI chatbot by simply prompting it. One adversarial session reportedly prompted the model more than 100,000 times across various non-English languages, collecting responses ostensibly to train a cheaper copycat. Google published the findings in what amounts to a quarterly self-assessment of threats to its own products that frames t…
Cybercriminals are attempting to uncover the “secrets” of advanced artificial intelligence (AI) models through thousands of structured questions. The global battle for dominance in the field of artificial intelligence is entering a new and more aggressive phase. The American company Google has confirmed that its advanced AI model, Gemini, has become the target of organized attacks by actors aiming to discover how the system works. According to r…
Coverage Details
Bias Distribution
- 57% of the sources lean Left
Factuality
To view factuality data please Upgrade to Premium
