Russian-speaking hackers have created artificial intelligence agents capable of autonomously identifying and exploiting cyber vulnerabilities, without human intervention, according to a report published Monday by Google's cyber threat intelligence division, cited by Politico.
The company says this is the first time it has observed the complete automation of the stages involved in discovering and exploiting a software vulnerability. The AI agents compare code before and after security updates are installed, detect weaknesses, and generate methods for compromising systems.
The attacks targeted websites in France, particularly in the media, entertainment, and education sectors. The technical documentation was written in Russian, but Google did not establish that the operation was coordinated by authorities in Moscow.
The development marks an important change, as complex attacks previously required experienced hackers for coordination and error correction. The risks increased after the launch, in April, of the specialized AI model Mythos, developed by Anthropic. The new capability increases the speed of attacks and reduces dependence on human specialists, complicating the protection of IT infrastructures.
Sources
Latest News
21:08
20:59
20:50
20:46
20:35
See more news