Iran used Claude to target US Navy in Middle East, Anthropic says
Iranian state-linked actors used the Claude AI model to research vulnerabilities in shipborne systems and target US Navy vessels. Houthi rebels also used the chatbot to develop code for hypersonic and ballistic missile guidance. Anthropic reports these activities did not result in operational weapons. These AI-driven efforts occur alongside a conflict costing the US military at least $33.4 billion, mostly for munitions. While OpenAI and Anthropic have linked their copyright policy shifts to data center spending, AI leaders recently attended a state dinner between Donald Trump and Xi Jinping.
Listen to Live Briefing
Real-time synthesized voice briefing · Live Feeds Desk
- ✓ Iranian state-linked actors used the Claude AI model to research vulnerabilities in shipborne systems and target US Navy vessels.
- ✓ Houthi rebels used the Claude chatbot to write code for hypersonic missile development and ballistic missile guidance.
- ✓ Anthropic states that these AI activities did not produce operational weapons.
- ✓ The US military has spent at least $33.4 billion on the conflict, primarily for munitions.
What changed
AI leaders attended a state dinner between Donald Trump and Xi Jinping, with one CEO dismissing the need for industry-wide coordination.
Live updates
-
Iran Used Claude AI to Target US Navy Vessels, Anthropic Reports
Iranian state-linked actors used the Claude AI model to research vulnerabilities in shipborne systems and target US Navy vessels. Houthi rebels also used the chatbot to develop code for hypersonic and ballistic missile guidance. Anthropic reports these activities did not result in operational weapons. These AI-driven efforts occur alongside a conflict costing the US military at least $33.4 billion, mostly for munitions. While OpenAI and Anthropic have linked their copyright policy shifts to data center spending, AI leaders recently attended a state dinner between Donald Trump and Xi Jinping.
Why it matters
The use of large language models for military targeting represents a shift in how state actors identify hardware vulnerabilities. This activity coincides with high US expenditure on munitions in the Middle East. The intersection of AI development and geopolitical tensions is further highlighted by the participation of AI CEOs in high-level US-China diplomacy.
What is confirmed
- Iranian state-linked actors used the Claude AI model to research vulnerabilities in shipborne systems and target US Navy vessels.
- Houthi rebels used the Claude chatbot to write code for hypersonic missile development and ballistic missile guidance.
- Anthropic states that these AI activities did not produce operational weapons.
- The US military has spent at least $33.4 billion on the conflict, primarily for munitions.
- OpenAI and Anthropic linked their rethink of copyright to data center spending.
Still unconfirmed
- One AI CEO dismissed the need for industry-wide coordination during a state dinner with Donald Trump and Xi Jinping.
What to watch next
- Evidence of operational weapons developed using LLMs
- Further disclosures from Anthropic regarding specific shipborne vulnerabilities identified by Iran
- Policy changes regarding AI industry coordination following the Trump-Xi state dinner
confidence 90%Sources used for this update (6)
- jen.jiji.com — Kazakhstan’s swimmer Sofia Spodarenko reaches Asian Games final
- jen.jiji.com — Tokayev thanks Putin for condolences over deaths of military personnel
- jen.jiji.com — Nor’easter threatens US East Coast as Tropical Storm Nolo nears Hawaii
- www.foxnews.com — AI leaders attend Trump-Xi state dinner as one CEO dismisses need for industry-wide coordination
- seekingalpha.com — Weekly Commentary: Too Big To Fail Redux
- jen.jiji.com — Kazakhstan’s Kholmogorov takes 2026 Asian Games canoe sprint bronze
-
Anthropic reports Iranian use of Claude to target US Navy
Iranian state-linked actors used the Claude AI model to research vulnerabilities in shipborne systems and target US Navy vessels. Houthi rebels also utilized the chatbot to write code for ballistic missile guidance and hypersonic missile development. Anthropic states these activities did not produce operational weapons. These efforts coincide with a conflict costing the US military at least $33.4 billion, primarily for munitions. Recently, OpenAI and Anthropic have linked their rethink of copyright to data center spending.
Why it matters
The use of generative AI for military research represents a shift in how state actors identify hardware vulnerabilities. This activity occurs amid high financial costs for the Pentagon and fluctuating diplomatic rhetoric regarding Iran.
What is confirmed
- Iranian state-linked actors used the Claude AI model to research shipborne system vulnerabilities and target US Navy vessels.
- Houthi rebels used the Claude chatbot to code ballistic missile guidance systems and attempt hypersonic missile development.
- Anthropic asserts that no operational weapons resulted from these AI-driven activities.
- The Pentagon reports a conflict with Iran costing the US military at least $33.4 billion, mostly for munitions.
Still unconfirmed
- OpenAI and Anthropic are tying a copyright rethink to data center spend.
What to watch next
- Updates on whether AI-generated code led to successful weapon deployments
- Pentagon reports on the efficacy of munitions used in the Iran conflict
confidence 90%Sources used for this update (2)
- www.afr.com — ASX to rise, Nasdaq extends tech rebound, oil trades below $US98
- www.capitalbrief.com — Trump floats Iran annihilation, then ‘very good’ talks
-
Iran and Houthi actors used Claude AI for US Navy targeting and missile coding
Anthropic's 2026 threat report reveals state-linked Iranian actors used the Claude AI model to research shipborne system vulnerabilities and target US Navy vessels. The report also found Houthi rebels used the chatbot to code ballistic missile guidance systems and attempt hypersonic missile development. Anthropic asserts that no operational weapons resulted from these activities. These AI-driven efforts occur as the Pentagon reports a conflict with Iran costing the US military at least $33.4 billion, mostly for munitions.
Why it matters
The use of commercial AI for military targeting highlights a new vulnerability in global security. This activity aligns with broader strategic interests in hypersonics and electronic warfare. The financial cost to the US military reflects the scale of the ongoing regional confrontation.
What is confirmed
- State-linked Iranian actors used the Claude AI model to target US Navy ships and research shipborne system vulnerabilities.
- Houthi rebels used the Claude chatbot to code ballistic missile guidance systems and attempt hypersonic missile development.
- Anthropic states no operational weapons were produced from these AI activities.
- The Pentagon reports the conflict with Iran has cost the US military at least $33.4 billion, primarily for munitions.
What to watch next
- Evidence of operational weapons developed via AI
- Updates to Anthropic safety filters regarding military applications
- Pentagon reports on the effectiveness of AI-targeted attacks
confidence 100%Sources used for this update (4)
- www.greenleft.org.au — Big data and the AUKUS war machine
- www.wvnews.com — Asian shares mostly rise, cheered by a US rally and easing oil prices
- sg.news.yahoo.com — Jayson Tatum Addresses Jaylen Brown Rift Rumors After Celtics Trade
- sg.news.yahoo.com — Lakers Tipped To Poach Bargain Defensive Wing In Trade With Wizards
-
Anthropic Report Links Iranian Actors to Claude AI for Military Targeting
Anthropic's 2026 threat report confirms state-linked Iranian actors used the Claude AI model to target US Navy ships and research shipborne system vulnerabilities. The report also identifies Houthi rebels using the chatbot to code ballistic missile guidance systems and attempt hypersonic missile development. While the AI assisted in military targeting and weapons programs, Anthropic states no operational weapons were produced. These activities occurred alongside a broader conflict with Iran that the Pentagon says has cost the US military at least $33.4 billion, primarily for munitions.
Why it matters
This incident highlights the use of commercial large language models for state-sponsored intelligence and weapons research. It occurs as the US manages high financial costs in the Middle East. The Pentagon is tracking these expenditures amid ongoing tensions with Iran and its proxies.
What is confirmed
- Anthropic's 2026 threat report identifies state-linked actors using Claude for military targeting and weapons programs.
- Iran used Claude to gather open-source intelligence to target US Navy ships in the Middle East.
- Houthi rebels used the chatbot to assist in coding ballistic missile guidance systems and researching hypersonic missiles.
- The Pentagon reports the Iran war has cost the US military at least $33.4 billion, mostly on munitions.
Still unconfirmed
- The AI was used for espionage and biological weapons research.
What to watch next
- Evidence of operational weapons derived from AI assistance
- Further Pentagon disclosures on non-military agency spending related to the Iran war
confidence 100%Sources used for this update (4)
- www.yahoo.com — White House backs Hegseth as new details come out about Iran war's cost
- www.yahoo.com — Iran-Linked Actors Used Claude to Target U.S. Warships and Develop Missiles
- www.theguardian.com — Democrats criticize Mike Johnson for sending lawmakers home for recess a day early – US politics live
- www.yahoo.com — Bucs Football | Morning Blend
-
Anthropic Report Links Iran, Houthi Rebels to AI Use for Naval Targeting
Anthropic disclosed in a 154-page security report that Iran used its Claude AI model to gather open-source intelligence to target US Navy ships in the Middle East and research vulnerabilities in shipborne systems. The company also found Houthi rebels in Yemen used the chatbot to assist in weapons development, including coding ballistic missile guidance systems and attempting to build hypersonic missiles. The report further identifies efforts to use the AI for espionage and biological weapons research.
Why it matters
This incident highlights the risk of large language models being repurposed for military intelligence and weapons engineering. It occurs amid heightened tensions in the Middle East involving Iranian-backed groups and US naval assets.
What is confirmed
- Iran used the Claude AI model to target US Navy forces in the Middle East.
- Houthi rebels in Yemen used Claude to develop ballistic missile guidance systems and hypersonic missiles.
- Anthropic released a 154-page security report detailing these AI abuses.
Still unconfirmed
- Users attempted to carry out biological weapons development using the AI.
What to watch next
- US Navy response to identified shipborne system vulnerabilities
- Anthropic updates to safety filters regarding weapons engineering
confidence 100%Sources used for this update (4)
- www.briefs.co — Saudis Shut East-West Pipeline After Drone Strikes
- worldisraelnews.com — Anthropic reveals Yemenis vibecoding ballistic-missile guidance, Iran using AI to target US Navy
- www.briefs.co — Chubu Electric chiefs set to quit after nuclear safety data scandal, NHK says
- www.singtaousa.com — Anthropic報告揭露 伊朗利用AI模型Claude鎖定中東美國海軍
-
Anthropic Says Iran Used Claude AI to Target US Navy
Anthropic reports that Iran and Houthi rebels utilized its Claude artificial intelligence model to conduct malicious cyber operations, build targeting material against United States Navy warships in the Middle East, and assist in weapons development. According to the company, users in Houthi-held Yemen also attempted to use the chatbot to develop advanced weapons, including building hypersonic missiles and coding ballistic missile guidance systems. The artificial intelligence firm disclosed these findings in a September 2026 report detailing various disruptions of bioweapons research and espionage attempts.
Why it matters
The misuse of commercial artificial intelligence models for military targeting and advanced weapons design highlights growing concerns over the proliferation of dual-use technologies by state and non-state actors. Anthropic's public disclosures reflect an increasing effort by technology companies to monitor and report how foreign actors exploit generative tools for espionage, cyber operations, and tactical military applications.
What is confirmed
- Anthropic reported that Iran and Houthi rebels used its Claude AI model to target U.S. Navy warships in the Middle East.
- An Iran-linked group used Claude to track U.S. Navy ships and build targeting material from publicly available data.
- Users in Houthi-held Yemen tried to develop advanced weapons and code ballistic missile guidance systems using the AI bot.
- Anthropic blocked possible efforts to build biological weapons during the same monitoring period.
Still unconfirmed
- Diesel prices in the United States hit six dollars a gallon as Houthi rebels took the Red Sea port city of Mokha.
What to watch next
- Further official disclosures from technology firms regarding state-sponsored artificial intelligence exploitation
- Additional regulatory responses or investigations concerning dual-use AI security controls
confidence 95%Sources used for this update (12)
- The New York Times — Anthropic Says It Blocked Possible Efforts to Build Biological Weapons
- Anthropic — Detecting and countering misuse of AI: September 2026
- www.yahoo.com — Diesel hits record $6 a gallon as Houthis take key Red Sea port city
- WSJ — Anthropic Says Iran Used Its American AI Model to Target U.S. Navy Warships
- Navy Times — Iran used Claude to target US Navy in Middle East, Anthropic says
- AP News — Users in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says
- Reuters — How Anthropic says Claude was used for weapons, spying and cyber operations
- Honolulu Star-Advertiser — Anthropic report details disruption of bioweapons research, cyber espionage on Claude
- axios.com — Anthropic report: 5 ways Claude was exploited for war, spying and repression
- www.briefs.co — Kalshi moves to launch US single-stock perps, seeks joint oversight
- interestingengineering.com — Anthropic says Iran-linked group used Claude to target U.S. Navy ships
- www.yahoo.com — Iran and Houthi rebels used Anthropic's Claude AI to target US warships and build hypersonic missiles — Houthi rebels also used the bot to code ballistic missile guidance systems
Community Sentiment: How do you assess this situation?
Voice your perspective · Real-time aggregated sentiment from the Live Feeds community