郭莉莉2026年9月15日北京的人民大会堂。中国情报部门负责人呼吁加强党对人工智能的管控，并加强对该行业的政府监管。 Kenny Holston/The New York TimesChina’s top spy chief has warned that artificial intelligence could pose a direct threat to the Chinese Communist Party’s hold on power, in what is the highest-level and most detailed articulation yet of how Beijing sees the technology’s security risks.中国最高情报官员警告称，人工智能可能对中国共产党的执政地位构成直接威胁，这是迄今为止北京方面就该技术安全风险所作出的最高层级、最详尽的表述。The unusually blunt assessment by Chen Yixin, the head of China’s Ministry of State Security, is in contrast to the warnings dominating the A.I. debate in the United States, which have focused on the possibility that A.I. could slip out of human control or pose a threat to humanity itself.中国国家安全部部长陈一新这一异常直白的评估与主导美国人工智能辩论的警告形成了鲜明对比，后者的焦点在于人工智能可能脱离人类控制，或对人类本身构成威胁。In an article published on Sunday in the state-run magazine China Cyberspace, Mr. Chen called for more party control over A.I. and stricter government oversight. This, he said, was necessary to protect political stability at home; to defend against increasingly sophisticated cyberattacks and misinformation campaigns from “hostile forces”; and to compete militarily with countries like the United States.在周日发表于官方杂志《中国网信》的文章中，陈一新呼吁党加强对人工智能的控制，并实施更严格的政府监管。他表示，这是为了维护国内政治稳定、抵御“敌对势力”日益复杂的网络攻击和虚假信息运动，以及在军事上与美国等国家竞争。Mr. Chen referred specifically to U.S. tech companies and their A.I. models — like Claude Mythos and GPT-5.5-Cyber — as examples of new threats to cybersecurity. He also described the danger that foreign intelligence agencies might use A.I. for “large-scale espionage” and attacks on China’s critical infrastructure.陈一新特别提到了美国科技公司及其人工智能模型——如Claude Mythos和GPT-5.5-Cyber——作为网络安全新威胁的例子。他还描述了外国情报机构可能利用人工智能造成境内数据“大规模外泄”，以及对中国关键基础设施发动攻击的危险。The warning from the powerful security agency makes clear that China sees foreign A.I. models as an active threat to China, according to Henry Gao, a law professor at Singapore Management University. That effectively draws a red line before talks between American and Chinese officials on A.I. safety risks, expected ahead of a meeting between President Trump and China’s leader, Xi Jinping, on Sept. 24 in Washington.新加坡管理大学法学教授高树超表示，这家强势安全机构的警告清楚地表明，中国将外国人工智能模型视为对中国的现实威胁。这实际上在美中官员就人工智能安全风险举行会谈之前划下了一道红线，预计该会谈将在特朗普总统与中国领导人习近平9月24日于华盛顿举行会晤之前进行。“Beijing is going into these talks viewing A.I. safety not as a shared humanitarian mission, but through an adversarial lens,” Mr. Gao said. It is making clear to Chinese tech firms and foreign rivals that “data sovereignty and political security will never be traded away for international safety accords,” he added.“北京带着这样一种视角参加这些会谈：不把人工智能安全视为一项共同的人道主义使命，而是通过对抗的视角来看待，”高树超表示。他还说，这向中国科技企业和外国竞争对手明确传达了一个信息：“绝不会为了国际安全协议而牺牲数据主权和政治安全。”Mr. Chen’s remarks are not the first time the security agency has spoken about the dangers of A.I., but they are the most comprehensive and signal Beijing’s growing concern about the industry. The security chief outlined several ways that A.I. could pose a threat to China:陈一新的言论并非该安全机构首次谈及人工智能的危险，但这些言论是最为全面的，并且标志着北京对该行业日益增长的担忧。这位安全主管概述了人工智能可能对中国构成威胁的几种方式：• Undermining the security of the Communist Party. The government’s top priority has long been maintaining the power of the ruling party. Mr. Chen’s article warns that those with “ulterior motives” could use A.I. to generate fake videos and audio to wage a “propaganda war” against China.• 损害共产党的安全。长期以来，政府的首要任务始终是维护执政党的权力。陈一新的文章警告称，那些“别有用心”的人可能会利用人工智能生成虚假音视频，对中国发动“舆论战”。• “They’re worried, among other things, of say a deepfake of a Chinese official saying something that wasn’t actually said, and that spreading rapidly and destabilizing everything,” said Kyle Chan, a fellow at the Brookings Institution who studies Chinese technology and industrial policy. “Someone thinks that Xi Jinping made some announcement which he didn’t actually. That could sow chaos.”• “他们担心的其中一件事是，比如出现一段中国官员的深度伪造视频，视频中官员说了一些实际上并未说过的话，并且该视频迅速传播，破坏了一切稳定，”布鲁金斯学会研究中国科技与产业政策的研究员凯尔·陈说。“如果有人以为习近平发布了某种实际上他并没有发布的公告。这可能会制造混乱。”• Attacks on China’s critical infrastructure. Mr. Chen, the spy chief, wrote that the technical and financial barriers to launching cyberattacks had been drastically lowered because of A.I. This, he said, posed “serious risks” to China’s information infrastructure. Last week, a small team of researchers in California demonstrated that risk when they disclosed a hacking tool that could take control of millions of users’ accounts on WeChat, China’s most widely used messaging app.• 针对中国关键基础设施的攻击。情报部门负责人陈一新写道，由于人工智能，发动网络攻击的技术和资金门槛已大幅降低。他表示，这对中国的信息基础设施构成了“严重风险”。上周，加利福尼亚州的一个小型研究团队展示了这种风险，当时他们披露了一款可以控制中国最广泛使用的即时通讯应用微信上成百上千万用户账户的黑客工具。• Leaks of sensitive information. Chinese users using foreign models could cause large-scale data leaks, the article noted. Anthropic, the company behind Claude and other A.I. models, said last week that the Chinese start-up Moonshot AI had routed queries by its users to Claude. Those queries included sensitive data, including video surveillance linked to the Chinese military, as well as proprietary information from high-profile Chinese technology companies.• 敏感信息泄露。文章指出，中国用户使用外国模型可能会导致大规模数据泄露。开发了Claude及其他人工智能模型的Anthropic公司上周表示，中国初创公司月之暗面曾将其用户的查询转发给Claude。这些查询包含敏感数据，包括与中国军方相关的视频监控信息，以及来自知名中国科技公司的专有信息。• More intense military competition. The U.S. military’s use of A.I. in the U.S.-Israeli war with Iran was an example of how warfare has been “fundamentally transformed,” Mr. Chen wrote. He says that whoever can harness the power of tools such as algorithms to achieve more precise sensing and targeting, “will gain the initiative on the battlefield.”• 更激烈的军事竞争。陈一新写道，美军在美以与伊朗的战争中使用人工智能，这是战争如何被“根本性变革”的例子。他说，谁能利用算法等工具的力量来实现更精确的感知和目标定位，“谁就能掌握战场主动权。”中国领导人习近平去年在北京参加阅兵式时的照片。习近平将人工智能视为一项战略性技术，认为它对中国与美国竞争的能力至关重要。The warning adds to a growing drumbeat of concern from Chinese officials about the risks of A.I. Under Mr. Xi, China has deemed A.I. a strategic technology that is critical to the country’s development and ability to compete with the United States. At the same time, officials have begun to issue warnings and new guidelines for regulating the industry.这一警告加剧了中国官员对人工智能风险越来越强烈的担忧。在习近平领导下，中国已将人工智能视为对国家发展及与美国竞争能力至关重要的战略技术。与此同时，官员们已开始发布警告并出台新的行业监管指南。Previous warnings from Chinese regulators have tended to focus on specific issues, such as when the Ministry of Industry and Information Technology said in July that it had detected a security risk in some versions of Anthropic’s coding tool, Claude Code. Regulators have also repeatedly warned domestic users about the risks of A.I. assistants like OpenClaw.以前中国监管机构的警告往往集中在具体问题上，例如工业和信息化部曾在7月份表示，在Anthropic的编程工具Claude Code的某些版本中检测到了安全风险。监管机构也曾多次警告国内用户警惕OpenClaw等人工智能助手的风险。“This is now a much more broad-based sort of warning, not about a specific threat from this one model or this one vulnerability, but a broader paradigm shift in how A.I. will create new risk for China,” said Mr. Chan of Brookings.“现在，这是一种更为广泛的警告，不是关于这个单一模型或这个单一漏洞带来的具体威胁，而是关于人工智能将如何为中国制造新风险的更广泛的范式转变，”布鲁金斯学会的凯尔·陈说。“It’s about taking these new powerful tools seriously and realizing there are a lot of vulnerabilities and China needs to move faster, and also use A.I. to patch those vulnerabilities, and defend against these new powerful cyber-actors,” he said.“这意味着要认真对待这些强大的新工具，并认识到存在许多漏洞，中国需要加快步伐，同时还要利用人工智能来修补这些漏洞，并抵御这些新型的强大网络行为体，”他说。Even as China’s security chief was articulating the country’s concerns about A.I., its Foreign Ministry was accusing U.S. companies of hyping the dangers of the technology.就在中国安全主管阐明该国对人工智能的担忧之际，中国外交部却在指责美国企业炒作该技术的危险性。That was in response to calls by some Western A.I. executives like Dario Amodei, the chief executive of Anthropic, for Washington to clamp down on Chinese companies.这是针对Anthropic首席执行官达里奥·阿莫迪等一些西方人工智能高管呼吁华盛顿对中企采取严厉措施所作出的回应。In an open letter over the weekend, Mr. Amodei said that democracies must keep their lead over China. He backed restrictions on selling powerful A.I. chips to the country, among other measures to slow China’s progress enough to widen America’s lead.在周末的一封公开信中，阿莫迪表示，民主国家必须保持对中国的领先地位。他支持限制向该国出售强大的人工智能芯片，并支持采取其他措施来放缓中国的进步，从而扩大美国的领先优势。Asked if China’s A.I. industry posed a threat to the United States, Guo Jiakun, a spokesman for China’s Foreign Ministry, said on Monday that such “threat narratives” and “malicious competition” would only make it harder for countries to work together on the regulation of A.I.当被问及中国的人工智能产业是否对美国构成威胁时，中国外交部发言人郭嘉昆在周一表示，这种“威胁叙事”和“恶性竞争”只会让各国更难以在人工智能的监管方面开展合作。In an editorial on Sunday, the Communist Party-run tabloid, the Global Times, said that China played a central role as one of the world’s main suppliers of A.I. tools. It described Mr. Amodei’s call for the United States to contain China’s development as an “unrealistic delusion.”在周日的一篇社论中，由共产党开办的小报《环球时报》表示，中国作为世界上主要的人工智能工具供应国之一发挥着核心作用。该报将阿莫迪呼吁美国遏制中国发展的做法描述为“根本无法实现的妄想”。郭莉莉(Lily Kuo)是《纽约时报》报道中国记者，常驻台北。翻译：纽约时报中文网点击查看本文英文版。