11.17.26 |

Russian speaking hackers unleash a new wave of AI cyber attacks, Google warns

Russian speaking hackers unleash a new wave of AI cyber attacks, Google warns

For years, cyber attacks have relied on hackers finding weaknesses in computer systems, stealing passwords or writing malicious code to break into networks. Artificial intelligence has made some of that work faster, but humans have generally remained in control. Now, Google says it has uncovered something different: a system capable of finding security flaws and exploiting them on its own, raising concerns about how quickly cybercrime could evolve.

The discovery was detailed in a November 16 report from Google’s Threat Intelligence Group and cybersecurity firm Mandiant. Investigators identified a system called exploit_agent that used multiple AI tools to examine software updates, detect weaknesses and automatically develop and execute attacks. The operation appeared to focus on websites in France’s media, entertainment and education sectors. Researchers also found stolen victim data on the system’s server, although the report did not establish how much of it had been obtained through the automated exploits.

Why Russia is being mentioned in the investigation

The connection to Russia comes from technical files discovered by Google’s researchers. Some contained instructions written in Russian, suggesting that the people operating the system may have been Russian-speaking. However, Google has not identified the hackers or linked the operation to the Russian government. That distinction matters as governments and cybersecurity companies continue investigating cyber espionage and attacks associated with multiple countries, including Russia, China, Iran and North Korea.

A new era of cyber attacks has arrived as Google uncovers AI hacking operation with Russian clues

A new era of cyber attacks has arrived as Google uncovers AI hacking operation with Russian clues

What makes this case different is the level of independence given to the technology. Imagine a company releasing a security update to fix a weakness in its website. Instead of waiting for a hacker to manually study that update, an AI system can compare the old and new versions, figure out what was vulnerable and attempt to break into computers that haven’t installed the fix. Google says it has now observed this discovery-and-exploitation process operating without a human directing each step.

What this new era of cyber attacks means for everyone

For businesses, the immediate concern is speed. Security teams already race to install updates before criminals exploit known weaknesses, and automated systems could make that window even shorter. Google also found that these tools are not necessarily difficult to detect: their rapid activity can leave unusual traffic and recognizable traces. But as AI improves, researchers expect attackers to become more efficient and potentially harder to spot.

For ordinary internet users, the findings do not mean that personal phones or bank accounts are suddenly being attacked by autonomous AI. The documented operation primarily involved vulnerabilities in web applications and organizational systems. Still, people can be affected when companies holding their information are compromised. Google’s warning is that cyber attacks are moving beyond hackers simply using AI as an assistant. In some cases, the technology is beginning to carry out the work itself.


Tags