In a new report, Google Threat Intelligence Group revealed how hackers are using artificial intelligence to carry out cyberattacks and fraud.
The most significant findings:
- AI can launch cyberattacks without human intervention. Google discovered a system capable of independently analyzing software updates, identifying vulnerabilities, developing exploits, and launching attacks.
- North Korean hackers are targeting cryptocurrency companies. Cybercriminals impersonated recruiters, lured employees to fraudulent websites, and modified Claude Code settings to automatically install a backdoor upon execution.
- AI-related bills reaching millions of dollars. Hackers compromised corporate cloud services and exploited victims' computing resources. In some cases, the resulting costs to affected companies reached $1 million.
- Armies of AI-powered bots. Researchers uncovered the development of a platform designed to manage more than 100,000 social media accounts. These bots are intended to imitate real human behavior and distribute political content.
Google's key conclusion: Cyberattacks are becoming faster, more scalable, and increasingly independent of human involvement. The next stage will involve more sophisticated, stealthier, and fully autonomous cyber intrusions.

