Google found attackers using AI to automatically analyze patches and execute exploits.
Google Threat Intelligence Group discovered that attackers are using a multi-agent framework to automatically analyze software patches, write, and execute exploit code.