China’s leader on Friday laid out his nation’s bid to shape the path of artificial intelligence, casting China as a champion of an open approach to the technology and a trusted ally of developing nations in advancing A.I.

上周五，中国领导人阐述了塑造人工智能发展方向的构想，将中国描绘为开放技术路线的捍卫者，以及帮助发展中国家推进人工智能发展的可靠伙伴。

The remarks by Xi Jinping highlighted the importance that China’s top political and governmental leadership place on artificial intelligence. Mr. Xi’s speech did not mention the United States by name, but the message was clear: China plans to compete as the world’s other A.I. superpower.

习近平的讲话凸显了中国最高党政领导层对人工智能的高度重视。讲话虽然没有点名美国，但传递的信息十分明确：中国志在作为全球另一人工智能超级大国参与竞争。

Speaking at an A.I. conference in Shanghai, Mr. Xi said that “A.I. development should not be a solo performance by a single country but a symphony of global collaboration.”

在上海举行的一场人工智能大会上，习近平表示：“人工智能发展不应该是某个国家的独奏，而应当是全球合作的交响。”

He described open-source A.I. technology, in which much of the software can be freely shared and modified, as a “rare and historical opportunity” to spread the benefits of A.I. globally. The technology, he said, must be shared by developing nations or raise the threat of “new historical injustices.”

他将开源人工智能技术——即软件的大部分代码可以自由共享和修改——描述为将人工智能红利惠及全球的“难得的历史性机遇”。他表示，这项技术必须与发展中国家共享，否则将带来“新的历史不公”。

Mr. Xi’s vision is assuredly one of national self-interest as well. China is generally regarded as trailing the United States in A.I., but it is catching up. That is particularly evident in the open-source software developed by China’s leading A.I. software companies like DeepSeek, Moonshot and Zhipu.

当然，习近平描绘的愿景同样服务于中国的国家利益。外界普遍认为，中国在人工智能领域仍落后于美国，但它正在迎头赶上。这一点在中国领先人工智能企业——如DeepSeek、月之暗面和智谱——开发的开源软件上体现得尤为明显。

Their A.I. systems are closing the gap with the leading American A.I. chatbots and question-answering programs, including OpenAI’s ChatGPT, Anthropic’s Claude and Google’s Gemini.

这些公司的人工智能系统正在缩小与美国领先的人工智能聊天机器人和问答程序间的差距，包括OpenAI的ChatGPT、Anthropic的Claude，以及谷歌的Gemini。

On Friday, a Chinese company, Moonshot, introduced a new model that it claimed performed as well as American models from Anthropic and OpenAI. The announcement of the model, Kimi K3, helped roil financial markets.

周五，中国的月之暗面发布了一款新模型，声称其性能可与Anthropic和OpenAI的美国模型相媲美。这款名为Kimi K3的模型一经发布，便助推了金融市场的波动。

The U.S. companies, which have spent many billions of dollars creating their frontier models, do not share their technology freely. They have also accused the Chinese of pilfering their technology. And China remains well behind in the most advanced A.I. chips, a market dominated by Nvidia.

这些美国公司为打造各自最先进的模型已投入数百亿美元，并不免费分享。它们还指责中国窃取技术。与此同时，中国在最先进的人工智能芯片领域仍远远落后——这一市场由英伟达主导。

Mr. Xi’s speech at the beginning of an A.I. conference in Shanghai came as China builds support for its World AI Cooperation Organization. The group was created a year ago, China said, partly to give developing nations a greater say in A.I. governance.

习近平在上海人工智能大会开幕式上的讲话正值中国为推动“世界人工智能合作组织”争取更多支持之际。该组织成立于一年前，中国表示，其目的之一是让发展中国家在人工智能治理中拥有更大的话语权。

In his address, Mr. Xi also said China planned to offer its A.I. technology and training to developing countries that were friendly to China. “We must uphold openness and win-win cooperation,” he said.

中国领导人在讲话中还表示，中国计划向对中国友好的发展中国家提供人工智能技术和相关培训。“坚持开放共赢，”他说道。

China’s international outreach is meant to supplement its A.I. ambitions at home. Mr. Xi has exhorted his country to “occupy the commanding heights” of technology, with A.I. at the forefront.

中国积极推动国际合作，、也是对其国内人工智能战略的补充。习近平曾敦促中国要“抢占科技制高点”，而人工智能正处于这一战略的最前沿。

But A.I., depending on how far and fast it progresses, poses a dilemma for the ruling Communist Party. How should it manage the rise of a technology that could one day be so disruptive that it could threaten its interests — and its grip on power?

然而，人工智能的发展——取决于其进步的速度和程度——也给执政的中国共产党带来了一个两难问题：如何管理一种未来可能具有巨大颠覆性的技术，使其不会最终威胁到党的利益，甚至危及其执政地位？

Chinese regulation of A.I. has varied in intensity over the years, depending on where the country assesses its strengths and weaknesses.

多年来，中国对人工智能监管力度时有变化，取决于官方对自身优势和短板的判断。

When the Chinese government was worried that it had fallen behind the United States in 2022, after the introduction of ChatGPT, it took a more hands-off approach that ultimately allowed ventures like DeepSeek and others to thrive.

2022年ChatGPT问世后，中国政府一度担忧自己落后于美国，于是采取了相对宽松的态度，这最终为DeepSeek等企业的蓬勃发展创造了空间。

At the same time, researchers have documented efforts by the Chinese government to monitor and manipulate public opinion in Hong Kong and Taiwan using companies with expertise in A.I. And the Philippines on Friday condemned as racist an A.I.-generated video posted by a Chinese state-run news site that depicted the Southeast Asian country as a karaoke-singing monkey doing the bidding of the United States and Japan.

与此同时，研究人员记录到，中国政府借助在人工智能领域具有专业能力的企业在香港和台湾监控并操纵舆论走向。菲律宾周五谴责中国一家官方媒体发布的一段AI生成视频涉嫌种族歧视——该视频将这个东南亚国家描绘成一只唱着卡拉OK、听命于美日两国的猴子。