Google Warns the ‘Era of Autonomous Attacks Has Arrived’

‘Instead of relying on manual effort to keep up, we will have to embrace agentic defense,’ the company said.

Google Warns the ‘Era of Autonomous Attacks Has Arrived’

TL;DR

  • Google warns of the emergence of autonomous AI attacks where AI continuously scans for and exploits software vulnerabilities without human intervention.
  • The speed and scale of these AI-driven attacks make it impossible to defend at human speed, necessitating the adoption of "agentic defense."
  • Threat actors are increasingly using large language models (LLMs) to automate the reverse engineering of patches and attack vulnerabilities.
  • Financial incentives include stealing LLM credentials and cloud resources to sell access to powerful AI models.
  • State-sponsored groups and other entities use LLMs for synthetic media, persona creation, and automated distribution.
  • The FBI has warned about criminals using ARTEX, an AI-powered tool, to carry out cyberattacks, lowering the technical barrier for attackers.
  • ARTEX has been used against various sectors globally, including financial services, government, and technology.
  • A China-based individual allegedly used ARTEX alongside Anthropic's Claude Code in attacks that stole personal information from South Korean financial institutions.
  • Recent outages at multiple banking institutions, technology companies, and government systems have occurred, with Homeland Security Secretary Markwayne Mullin suggesting AI could be a cause.