AI Agents Power Lone Hacker's Website Blitz

A single hacker, reportedly Chinese-speaking, has successfully breached over 100 websites utilizing AI agents. This individual leveraged open-source AI platforms like Cairn to identify system weaknesses and execute cyberattacks on businesses across the United States, Canada, Australia, and Japan. The AI agents were capable of breaching sites in under a day, with the entire operation costing around $5,000 for token usage.

The AI agents performed complex tasks, including scanning digital records and crafting exploits to trick servers into executing malicious commands. Despite the sophisticated nature of the attacks, the prompts required were minimal, often as simple as 'fetch credit cards.' The AI models primarily used were China's DeepSeek and Anthropic's Claude Code. The creator of Cairn stated the platform was not designed for hacking, but acknowledged its potential misuse.