Anthropic’s latest threat intelligence report reveals that malicious actors are increasingly using artificial intelligence to automate and scale cyberattacks, surveillance, propaganda, and weapons development across global operations.

Key Points

• Anthropic identified malicious activity between December 2025 and August 2026, including cyber operations, fraud, and weapons research.
• AI is being used to automate entire attack chains, from reconnaissance and credential harvesting to malware development and data exfiltration.
• State-sponsored groups and criminals are leveraging AI to conduct mass surveillance, influence campaigns, and large-scale consumer fraud.
• The report documents AI assistance in developing guided rockets, ballistic missiles, and autonomous drone swarms by actors in China, Russia, and Yemen.
• Several Chinese AI companies, including Alibaba and Zhipu, were caught illicitly extracting model capabilities and rerouting user data through fraudulent accounts.

Why it Matters

The integration of AI into malicious operations significantly lowers the barrier to entry for sophisticated attacks by reducing the need for large, specialized human workforces. This shift fundamentally alters the economics of cybercrime and national security, as AI allows small teams to execute complex, large-scale campaigns at machine speed.