🕒 Created

Anthropic reports Iranian actors used Claude AI to target U.S. Navy forces and develop weapons systems

The administration announced that Iranian threat actors utilized Anthropic's Claude AI software to identify and track U.S. Navy positions in the Middle East. According to a report from Anthropic, these actors used the software to compile targeting handbooks and research vulnerabilities in shipboard systems, including maritime VSAT terminals and Cisco communications equipment. The Iranian actors also used the AI to develop an automated identity-profiling harness to target hundreds of Israeli governmental and non-governmental individuals, as well as Jewish diaspora organizations. Furthermore, the report detailed how a northern-Yemeni group used Claude to develop guidance software for high-end weapons, such as multi-stage ballistic missiles. Anthropic stated it has banned the accounts involved, created safeguards to prevent similar activity, and shared the intelligence with government agencies. While the administration's report highlights these capabilities, the AI company also noted that Iran used the technology for domestic surveillance and propaganda. Iranian organizations used the AI to build operational manuals and transform official government material into targeted messaging. Anthropic also noted that the technology was used to run campaigns tied to 'Jihad al-Tabyin' to shape public opinion both at home and abroad.

Sources

Paywall and unreadable sources