report<\/u><\/a><\/span>, which was released on Thursday, described a \u201ccell of threat actors based in northern Yemen\u201d trying to use the company\u2019s Claude AI to create a ballistic missile guidance system.<\/p>\n\u201cThe actors used Claude Code in place of human software engineers to develop the guidance, navigation and control software that steers and stabilizes a flying vehicle,\u201d the company wrote. \u201cFor example, they used Claude to integrate an open-source autopilot onto a phone-class flight computer, writing the control and position estimation software, tuning the control settings, running a firmware build pipeline and performing a flight simulation.\u201d<\/p>\n
Using large language models like Claude to write functional software code without knowing how to write a programming language is commonly known as \u201cvibecoding\u201d and has been one of the revolutionary developments driving Anthropic\u2019s rise to what is likely to be the largest initial public offering in history later this year.<\/p>\n
The report did not name the Yemeni \u201cthreat actors,\u201d but its description of their development of \u201ca multi-stage ballistic missile with a stated range goal above 2,000 kilometers\u201d likely refers to the Iran-backed Houthis, who have have used ballistic missiles to attack Israel, Saudi Arabia and other targets around the region.<\/p>\n
Anthropic said that it \u201cdisrupted\u201d the Yemeni cell and informed \u201cpublic- and private-sector partners to mitigate risks.\u201d But the group used Claude to make progress in its weapons development programs, Anthropic said.<\/p>\n
\u201cWe do not have evidence the actors succeeded in fielding an operational device, but they did test-fire a guided rocket,\u201d the company said. \u201cThis field test appears to have failed: within hours, the actors returned to Claude to work out why it failed.\u201d<\/p>\n
The company says it banned the accounts, but \u201cwe have evidence that the actors had already built an offline simulation toolkit that does not rely on Claude or other engineering computing environments\u201d to carry out the work.<\/p>\n
In another case study, Anthropic said that \u201can Iran-nexus threat actor\u201d used Claude to collect open-source data to generate targeting recommendations against U.S. Navy facilities in the Middle East.<\/p>\n
\u201cThe compiled material included a roster of U.S. personnel scraped from captions on public military photographs, publicly accessible ship and aircraft transponder identifiers, commercial satellite-imagery query scripts and an inventory of public websites that exposed U.S. naval movements,\u201d the company said.<\/p>\n
The 154-page security report from the AI company also warned of suspected attempts to carry out biological weapons development and to use its models to support a propaganda campaign from Iran\u2019s Ministry of Culture and Islamic Guidance to wage \u201cexplanatory jihad.\u201d<\/p>\n
\u201cWe\u2019re publishing this work, because we believe we have a responsibility to disclose malicious misuse of our services,\u201d the company said. \u201cAs models become increasingly capable, their risks will increase, unless AI developers and society\u2019s defenders act to make them safer.\u201d<\/p>\n<\/div>\n<\/div>\n
\n
\n
\n
\n
Andrew Bernard is the Washington correspondent for JNS.org.<\/em><\/p>\n<\/div>\n<\/div>\n<\/div>\n
![\"\"](\"https:\/\/www.crescentcityjewishnews.com\/blog\/wp-content\/uploads\/2026\/09\/Claude_Anthropic-300x166.png\")
![\"Share](\"data:image\/gif;base64,R0lGODlhAQABAAAAACH5BAEKAAEALAAAAAABAAEAAAICTAEAOw==\")
![\"Share](\"https:\/\/static.hupso.com\/share\/buttons\/share-medium.png\")
![\"\"](\"https:\/\/www.crescentcityjewishnews.com\/blog\/wp-content\/uploads\/2026\/09\/Claude_Anthropic.png\")