North Korea says it tested an AI-enabled Zam-2.0 torpedo, but independent verification of the weapon’s capabilities has not been reported.
North Korea said on Saturday, October 17, that it had successfully tested a new torpedo system using artificial intelligence for target detection and guidance. The Korean Central News Agency identified the weapon as Zam-2.0 and described it as a high-speed, precision-guided system. The test reportedly took place on Thursday, October 15.
KCNA said the trial checked the weapon’s propulsion systems, which it said had been developed through more than 180 tests over five years. It did not explain what AI techniques were used or provide technical details about how the system detects targets or guides the torpedo. South Korea’s military had not immediately commented on the announcement when the report was published.
The claims should be treated cautiously: the reported capabilities come from North Korean state media, and the announcement offers no independent evidence that the torpedo can identify targets autonomously. Experts cited in the reporting questioned whether North Korean weapons have advanced capabilities such as autonomous target designation, while noting the country may be improving its guidance software.
The announcement comes amid a wider series of North Korean weapons tests, including recent submarine-launched missile activity. It also follows earlier North Korean claims that AI was used in missile systems. Those claims, like the Zam-2.0 announcement, have not established what functions the AI performs or how reliably they work.
For defense and AI-policy observers, the development is notable as a public claim of AI in an underwater weapon—not as confirmation of a proven autonomous targeting capability. The available report provides no evidence that Zam-2.0 was deployed or used operationally. Further information from independent sources or technical assessments would be needed to evaluate the system’s actual performance.