Iran and Houthi Rebels Exploit Anthropic Claude AI to Target U.S. Warships and Build Hypersonic Missiles

Iran-Linked Actors Target U.S. Naval Forces Using Anthropic Claude AI

An Iran-linked threat actor used Anthropic Claude AI to support military reconnaissance and develop targeting recommendations against U.S. naval forces in the Middle East, according to Anthropic's September 2026 threat report1

1

. The perpetrator combined publicly available ship and aircraft transponder identifiers with commercial satellite imagery and information on U.S. naval movements3

3

. The actor extracted names of U.S. military personnel from captions of publicly available military photographs and researched potential vulnerabilities in communications equipment used aboard ships, including known flaws affecting Cobham Sailor VSAT terminals, Cisco communications equipment, and Schneider Electric EcoStruxure systems1

1

. Anthropic banned the account, introduced additional detection mechanisms, and shared its findings with government authorities1

1

.

This AI misuse demonstrates how adversaries exploit American-made technology against U.S. interests. Access to Claude is blocked in Iran, but adversaries easily create accounts using VPNs and foreign phone numbers3

3

. The findings align with how Iran has been using technology to support its strategic objectives in geopolitical conflicts with the U.S. and Israel3

3

.

Houthi Rebels Develop Advanced Weapons with AI

A cell in northern Yemen controlled by Houthi rebels used Claude Code to support three weapons programs: a guided rocket using a phone-class flight computer that assists terminal guidance, a multistage ballistic missile targeting a range of more than 2,000 kilometers, and an R2000 missile family that included a hypersonic glide vehicle variant1

1

. The group used Claude to develop guidance, navigation, and control software; integrate an open-source autopilot with a phone-class flight computer; write control and position-estimation code; tune parameters; build firmware; and run flight simulations1

1

. Essentially, the group used multiple Claude instances instead of a group of software engineers for coding, code review, research, and simulation1

1

.

Anthropics identified that users did not succeed in fielding an operational device but did carry out a failed test of a guided rocket2

2

. The company knows this because the users returned to Claude chatbot to find out why it failed2

2

. Anthropic confirmed it had evidence that before the account bans, users had already built an offline simulation tool kit that doesn't use Claude or any other computing platforms2

2

.

Trevor Ball, a weapons analyst at Armament Research Services, noted that while Houthis might be exploring hypersonic missiles by asking Claude, they lack the production or technical capabilities to actually build them2

2

. He suggested they are probably trying to develop their own capabilities to become less reliant on Iranian shipments of weapons and components2

2

. The revelation comes as Iran-backed militants have launched a wide-ranging offensive across Yemen towards the strategic Bab al-Mandab strait, which connects Europe to Asia4

4

.

AI Makes Targeting U.S. Military Operations Easier Than Ever

Experts warn that AI-enabled weapons development and AI for military operations could become increasingly common as America's adversaries turn to AI to boost military capabilities. "The new threat is just that it is now easier than ever before for adversaries to be able to understand what the U.S. military is doing," said Sam Bresnick, a research fellow at Georgetown's Center for Security and Emerging Technology3

3

. "It is getting much harder for the U.S. military to cloak or to obscure its activities."3

3

Nikita Shah, a senior fellow with the Intelligence, National Security, and Technology program at the Center for Strategic and International Studies, explained that adversaries are using AI to enhance tactics they already employ. "It just makes it much more effective, much more efficient, much faster as to how the models can help them carry out those different attack methods,"3

3

she said. AI can help refine the targeting of missiles and drones, increase situational awareness of the battlefield, and be used for monitoring3

3

.

The increasingly advanced AI models are allowing other countries to catch up to the U.S. advantage in situational awareness—something the Pentagon is already working to combat with countermeasures3

3

. A July report released by threat intelligence provider Recorded Future found that AI likely accelerated Iran's existing cyber and military capabilities in kinetic warfare3

3

.

Iran Deploys AI for Propaganda and Domestic Surveillance

Beyond military applications, Iran leveraged AI for propaganda and domestic surveillance operations. Anthropic found three Iranian state-aligned accounts using Claude for public influence campaigns as part of cognitive warfare programs3

3

. Iranian institutions produce propaganda as both a religious and strategic duty in what they call "soft war"5

5

. In one campaign involving Iran's war with the U.S. and Israel, actors falsely attributed claims to research institutions like think tanks Brookings and RAND to create a sense of credibility3

3

.

Two Iranian actors used the company's model to surveil and profile 6,388 Iranians in one year, analyzing 155,216 tweets1

1

. Another group used the model as an engineering pipeline to develop domestic surveillance tools, including the production-deployed "al-Najm al-thāqib" Firefox extension designed to mass-harvest user identities across major social platforms1

1

. These operations targeted Iranian minorities, dissidents, and opponents of the Iranian regime abroad using open-source intelligence5

5

.

Broader Implications and Response

Anthropics 154-page threat report, which covered findings from December to August, described cases ranging from state-sponsored groups spreading propaganda to unnamed actors researching how to make biological weapons more deadly2

2

. Iran-linked actors and Houthis are not the only entities using Anthropic's AI technologies for weapon development—China and Russia are also actively using Claude for their military programs1

1

.

Anthropics banned 16 Claude accounts associated with Iranian paramilitary and domestic security agencies, though that does not mean it has banned all of them1

1

. The company shared its findings with private and public partners2

2

. "We hope that the findings in this report will help other developers recognize similar patterns on their own platforms, give governments and civil society a clearer view of how emerging threats take shape, and strengthen collective defenses," Anthropic stated5

5

.

The report comes as debate over AI dangers rages among policymakers and AI company employees. A top scientist at Anthropic, Evan Hubinger, quit last week citing fears that AI companies were developing uncontrollable systems, writing "We really do earnestly believe AI could kill all humans!"5

5

This occurred even as Anthropic refused the Pentagon's February request to remove safety guardrails regarding mass domestic surveillance and fully autonomous weapons5

5

. Watch for increased scrutiny on AI safety protocols and potential regulatory actions as cyber campaigns and AI misuse by adversaries continue to evolve.