North Korea says it has test-fired a new precision-guided torpedo that uses artificial intelligence in its target detection and guidance systems. State news agency KCNA reported on Saturday that the Zam-2.0, described as a “super-speed” torpedo, was fired in an underwater test on Thursday, October 15, to confirm a technical upgrade, according to Yonhap’s account of the announcement.
Almost everything known about the weapon comes from Pyongyang itself. KCNA said the test proved core technology for propulsion efficiency and rocket propulsion devices, and claimed the programme has conducted more than 180 related test rounds over five years. It disclosed no speed, range or test location. The test was reportedly run by the Academy of Defense Sciences with the Missile General Bureau and the Mine and Torpedo Research Bureau, and Kim Jong Un did not attend. No independent body has confirmed the test took place as described, and nothing outside North Korean state media verifies that the torpedo’s AI does what is claimed.
Analysts in Seoul urge caution on the label as much as the launch. Experts cited by Yonhap and the Korea Times say the “super-speed” wording and the mention of rocket propulsion suggest the weapon could be a supercavitating torpedo — a class of weapon that travels inside a bubble of gas, and at whose speeds reliable guidance is a known unsolved engineering problem. Lim Eul-chul of Kyungnam University said the test appears aimed at showcasing a torpedo that is hard to detect and hard to evade, aimed at major South Korean surface ships; Shin Jong-woo of the Korea Defense and Security Forum said it is likely intended for the North’s expanding submarine force.
Why it matters: the claim lands in a week of naval signalling — two days before the torpedo announcement, Pyongyang launched two types of strategic cruise missiles from what was presumed to be a 3,000-ton submarine, and it has recently commissioned the 5,000-ton destroyers Choe Hyon and Kang Kon. Whether the Zam-2.0 works as advertised is unproven; that AI is now the vocabulary of military prestige is not. The same technology is being argued over in economic terms — Anthropic’s scenarios for what AI does to GDP by 2030 — and in courtrooms, where eight US states have sued OpenAI over AI safety. GEroundtable will treat the Zam-2.0 as what it is for now: a claim, worth watching, not yet a capability.
Source: Yonhap (relaying KCNA); Korea Times
0 Comments