Google’s Threat Intelligence Group (GTIG) has issued a report highlighting increasing risks associated with artificial intelligence (AI) in cyberattacks. The report reveals that state-sponsored hackers are leveraging AI tools like Google’s Gemini to enhance their operations, including model extraction for intellectual property theft and advanced phishing scams. The report identifies notable actors from North Korea, Iran, China, and Russia using large language models (LLMs) for nuanced target profiling and generating hyper-personalized phishing lures, effectively bypassing traditional detection methods. GTIG warns of a concerning trend towards agentic AI, which enables malware development with autonomy. Google is actively countering these emerging threats by publishing regular threat intelligence reports and improving Gemini’s security to mitigate misuse. Despite the escalating AI threat landscape, the report emphasizes that, so far, there have been no breakthrough capabilities in malicious AI use. Google continues its mission to identify and neutralize these risks proactively.
Source link
