AI becoming more dangerous in the hands of cybercriminals, Anthropic warns
The AI company Anthropic recently stopped a cybercriminal who used Claude Code to write phishing emails, create malicious code, and circumvent safety filters.
The latest edition of Anthropic’s Threat Intelligence Report shows that hackers, cybercriminals, and other threat actors have adapted their methods to exploit AI’s most advanced capabilities.
AI models are being used to perform sophisticated cyberattacks. In addition, AI technology has lowered the bar for engaging in cybercriminal activities. Even criminals with few technical and coding skills are now using AI to pull off ransomware attacks or phishing schemes.
Lastly, scammers and fraudsters have embedded AI throughout all stages of their operations, including profiling victims, analyzing stolen data, and creating false identities to expand their reach to more potential targets.
In its threat report, Anthropic discusses several recent examples of Claude being misused, including a large-scale extortion operation using Claude Code, a fraudulent employment scheme from North Korea, and the sale of AI-generated ransomware by a cybercriminal with only basic coding skills.
In the first case, the threat actor used Claude Code to automate reconnaissance, harvesting victims’ credentials, and penetrating networks.
“Claude was allowed to make both tactical and strategic decisions, such as deciding which data to exfiltrate and how to craft psychologically targeted extortion demands. Claude analyzed the exfiltrated financial data to determine appropriate ransom amounts and generated visually alarming ransom notes that were displayed on victim machines,” Anthropic states.
According to the researchers, this incident marks an evolution in AI-assisted cybercrime.
“Agentic AI tools are now being used to provide both technical advice and active operational support for attacks that would otherwise have required a team of operators. This makes defense and enforcement increasingly difficult, since these tools can adapt to defensive measures,” they say.
These kinds of sophisticated cyberattacks will only become more common as AI-assisted tools become more refined and available. Anthropic believes that the best solution is to improve methods for detecting and mitigating the harmful uses of its models, promising to conduct further research in this area.
