Anthropic published its latest threat intelligence report on September 11, 2026, detailing how multiple actors used its Claude AI models for activities spanning weapons development, cyber operations, surveillance, and fraud. The report covers incidents between December 2025 and August 2026 and marks the company's first comprehensive disclosure of this kind for the year.
One of the most striking revelations involves a China-based actor that used Claude to develop an electronic-warfare and air-defense suppression software suite. The software ranked targets and modeled radar jamming scenarios. During the project, the actor modified a simulation to include 12 targets in Taiwan — including early-warning radar, Patriot and Tien Kung missile batteries, and air bases. Anthropic described this as a clear case of AI being used to support military planning against a specific geopolitical target.
The report also identified a Russia-based cyber espionage campaign consistent with the Midnight Blizzard actor. This group used Claude to build a system that automatically detected when its malware was flagged by security defenses and rewrote the code until it evaded detection — a significant escalation in how AI can be used to automate and improve malware development.
On the biological weapons front, Anthropic said it blocked five separate cases where actors attempted to use Claude in ways that could support biological weapons development. The company stressed that the same information can be used for both weaponization and vaccine development, making this one of the most difficult dual-use challenges in frontier AI safety.
Additionally, the report documented six cases where Claude was used to develop software for conventional weapons — including firearms, missiles, armed drones, bombs, and other munitions, along with the targeting and control systems that operate them. An Iranian propaganda institution was also identified using Claude for influence operations.
Anthropic said it disrupted all identified malicious uses of its models. The report comes amid growing calls for AI regulation: U.S. Senator Bernie Sanders introduced legislation to ban AI superintelligence and temporarily pause advanced AI development, while OpenAI chief scientist Jakub Pachocki called for voluntary industry slowdowns until adequate safeguards are in place.