Breaking
Markets
EUR/USD1.1770 0.04%GBP/USD1.3607 0.03%USD/JPY157.15 0.01%USD/CHF0.7782 0.09%AUD/USD0.7245 0.06%USD/CAD1.3673 0.04%USD/CNY6.8093 0.19%USD/INR95.37 0.06%USD/BRL4.8968 0.10%USD/ZAR16.43 0.08%USD/TRY45.39 0.02%Gold$4,736.20BTC$81,341 0.36%ETH$2,323 1.65%SOL$96.78 0.60%
North America

Google Says It Thwarted Hacker Group's AI-Driven 'Mass Exploitation' Plot

Google said it blocked a hacker group trying to use its Gemini model to automate vulnerability discovery for a 'mass exploitation event.' The group sought to scan open-source repositories at scale. The disclosure renews concerns about generative AI being repurposed for offensive cyber operations.

Rows of servers in a darkened data center
Photo: Brett Sayles / Pexels
CNBC Top News1 h agoGOOGL MSFT

Google's Threat Intelligence Group disclosed that an unnamed hacker group tried to use Gemini to auto-scan open-source repositories for vulnerabilities. The goal was a 'mass exploitation event' hitting multiple projects at once.

Google says it detected the prompts, blocked the accounts and prevented offensive use of the model. The company is publishing the case to alert industry peers and government agencies.

Regulators in the US, UK and EU are preparing to tighten requirements on model providers to detect and report malicious use. Microsoft, Anthropic and OpenAI have announced similar takedowns in recent months.

AITechRegulationGOOGLMSFTNorth AmericaCNBC Top News
This article is an AI-curated summary of the original story published by CNBC Top News. The illustration is a stock photo by Brett Sayles from Pexels and is not from the original story.

More from North America