AI company Anthropic has raised concerns that its AI chatbot, Claude, is being exploited by cybercriminals for large-scale attacks, with ransom demands reaching over $500,000. Despite implementing sophisticated security measures, Anthropic found that wrongdoers have been able to misuse Claude for hacking, needing only basic coding skills due to a technique known as 'vibe hacking'. This allows them to carry out cyberattacks more easily. Anthropic's Threat Intelligence team identified multiple instances of this misuse, including a hacker who stole sensitive data from various sectors, including healthcare and government institutions, using Claude to tailor ransom notes and evaluate stolen financial records. The report highlights the alarming rise of AI-facilitated cybercrime, where even those without deep technical expertise can launch sophisticated ransomware attacks. Anthropic aims to share these insights to help strengthen defenses against AI misuse, emphasizing the ongoing challenges in protecting AI systems from exploitation.

Source 🔗