US AI firm Anthropic, creator of the chatbot Claude, has revealed that hackers and scammers are misusing AI for sophisticated crimes.

Cyberattacks: Hackers used Claude to write code that breached at least 17 organizations, including government offices. The AI helped in planning attacks, choosing which data to steal, and even suggesting ransom amounts, showing how AI can make both tactical and strategic decisions in cybercrime.

North Korean job scams: AI was also used to create fake profiles and secure remote jobs at top US companies. Once employed, the scammers used AI to translate messages, write code, and bypass cultural or technical barriers. This represents a new phase in employment-related fraud, making it harder for companies to detect.

Experts warn: AI is speeding up the time it takes to exploit cybersecurity vulnerabilities, which means organizations must shift from reactive to proactive security measures. While traditional methods like phishing are still common, AI amplifies attacks and increases efficiency.

Key takeaway: Organizations should treat AI as sensitive information, just like personal or financial data. Anthropic has disrupted attacks, reported them to authorities, and strengthened detection tools, but the incident highlights the growing risks of AI misuse in cybercrime.