Afghanistan, Nov. 18 -- Google has warned that cybercriminals are using autonomous artificial intelligence (AI) systems to identify software weaknesses and launch cyberattacks with little or no human involvement.
In a report published on November 16, the Google Threat Intelligence Group (GTIG) said attackers were increasingly using autonomous large language model (LLM) frameworks to automate several stages of cyber operations. The report covers findings from the third quarter of 2026.
According to Google, these systems can continuously examine software updates, compare code before and after security patches, identify weaknesses and develop working exploits. Some systems can also launch attacks automatically.
The company said it had ident...