Latest Anthropic horror story chills with tales of kamikaze drone swarms and bioweapons research

Anthropic, the artificial intelligence model provider, has released a chilling report detailing the misuse of its Claude models by cybercriminals and state-sponsored hackers. The report covers incidents from December 2025 to August 2026, detailing how adversaries have exploited Claude to conduct cyberattacks, build kamikaze drone swarms, carry out mass surveillance, and attempt to develop a more dangerous form of a deadly mosquito-borne virus.

Chinese spies have long been known to use Claude to automate digital intrusions and steal sensitive data from critical organizations. However, the latest report reveals that other actors, such as Russian freelancers and ShinyHunters, are also leveraging Claude for malicious purposes. These cybercriminals have been known to target various organizations, including embassies, think tanks, defense-industrial companies, and government agencies across multiple regions.

In addition to cyber threats, Anthropic's report highlights the potential biological misuse of Claude. The AI model has been used to support research related to chikungunya virus and highly pathogenic avian influenza. While the former research could potentially aid in developing better vaccines, the latter could have severe consequences, such as increasing disease severity, confusing diagnosis, and hindering treatment.

Furthermore, conventional weapons development has also been a concern. Through Claude, users in unsupported regions have been able to develop software for weapons such as firearms, missiles, armed drones, bombs, and targeting and control systems. This misuse includes instances in China, Russia, and Yemen, where Claude was used to develop guidance, navigation, and control software, as well as develop a Chinese-language specification for an anti-torpedo fire control system.

Anthropic's report emphasizes that while they have taken steps to block some of these malicious activities, they acknowledge that not all requests are successfully thwarted. The company remains committed to recognizing similar patterns on their platforms, providing governments and civil society with a clearer view of emerging threats, and strengthening collective defenses against AI misuse.