{
  "id": 5326088,
  "title": "Chinese military’s AI attack chain lags foreign models: researchers",
  "url": "https://urgent.news/2026/09/03/chinese-militarys-ai-attack-chain-lags-foreign-models-researchers-5326088",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-03T12:00:06.000Z",
  "source": {
    "name": "Reuters Business via SCMP",
    "slug": "reuters-business-via-scmp",
    "url": "https://www.scmp.com/news/china/military/article/3366256/chinese-militarys-ai-attack-chain-lags-foreign-models-researchers"
  },
  "original_language": "en",
  "account": "A study by Chinese researchers reveals that the Chinese military's AI capabilities are lagging behind their foreign counterparts, focusing on support functions instead of combat. The findings, published in the journal Defence Industry Conversion in July, emphasize the military's AI applications in areas such as intelligent question-answering and content generation. These functions still trail foreign large language models significantly and the offensive uses in actual combat remain in an exploratory stage with limited research. The research coincides with AI becoming a core element of the strategic competition between the US and China. The US is aiming to establish an \"AI-first\" warfighting force, while China is pursuing a military AI agenda to integrate emerging technologies into the People's Liberation Army. A notable example is the Chinese air force using an AI-enabled system to develop battle plans for operations involving over 100 units, reported CCTV in August. The journal's paper heavily relied on an American military analysis from Military Review, which examined AI's integration with cyber and information operations in US military operations. The authors noted three key capabilities of the US military's AI usage: expanding AI military applications from information processing to direct attacks, causing physical facility damage and cognitive effects in cyberattacks, and integrating AI into military command and decision-making. They urged China to develop an autonomous and controllable AI attack chain, capable of identifying and striking battlefield targets, creating code to attack enemy networks and command chains, and executing strikes in the cognitive domain. However, the study also pointed out a lack of systematically trained personnel in AI and big data within China's military, highlighting a potential gap between the country's rapid civilian AI advancements and their application on the battlefield. In 2017, the State Council set a goal for China to become the world's leading AI power by 2030, and the PLA started developing disciplines in AI and big data around the same time. Despite these efforts, the paper claims that a sizable professional cadre is still lacking, and specialists recruited from industry lack military background and are limited in operational settings. The researchers suggested sending serving personnel to civilian universities for specialized training and involving civilian experts in military exercises and routine combat readiness duties to bridge the gap.",
  "summary": "China’s military AI models trail their foreign equivalents in performance, concentrating on support functions rather than combat, according to a study by Chinese researchers. The findings appeared in a paper by researchers affiliated with the Armed Police Force Shandong Corps and the Engineering University of People’s Armed Police, published in the journal Defence Industry Conversion in China in…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 2,
    "also_reported_by": [
      {
        "outlet": "South China Morning Post",
        "title": "Chinese military’s AI attack chain lags foreign models: researchers",
        "url": "https://urgent.news/2026/09/03/chinese-militarys-ai-attack-chain-lags-foreign-models-researchers",
        "published": "2026-09-03T12:00:06.000Z"
      }
    ]
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}