- Hackers are using LLMs to write better phishing emails, scan targets’ files, and find overlooked bugs. - Recently Hackers have misused Anthropic’s AI for cyber-attacks, theft, and extortion. Hackers used Claude to write malicious code and plan attacks on at least 17 organisations (incl. govt bodies). - North Korean actors used Claude to create fake profiles and apply for remote jobs at US Fortune 500 tech firms. - AI helped attackers decide which data to steal, how to craft ransom demands, and even suggest ransom amounts. - Cyber firms (Google, CrowdStrike) are deploying AI to detect vulnerabilities - Governments and companies are racing to ensure AI helps defenders more than attackers. But as “agentic AI” grows, the risk of autonomous hacking escalates. - This incident highlights the dual-use nature of AI: while powerful for productivity, it can also amplify cybercrime and fraud (especially with agentic AI that can act autonomously). https://www.anthropic.com/news/detecting-countering-misuse-aug-2025 https://www.nbcnews.com/tech/security/era-ai-hacking-arrived-rcna224282