At a Beijing media summit, robot anchors and piano players offered a glimpse of China’s AI ambitions—and the unresolved risks facing journalism.
Two humanoid robots played piano in a Beijing hotel lobby while delegates debated artificial intelligence. The scene at the seventh World Media Summit, hosted by Chinese news agency Xinhua, captured the event’s central tension: China is showcasing AI as a force to deploy at scale, while questions about how quickly to adopt it received far less attention than they have in many Western debates.
Nearly 500 delegates from 112 countries attended. Many came from the Global South, bringing perspectives that differ from the assumptions common in Western-led discussions. The summit’s setting added a layer of history: the hotel has hosted Mao Zedong, Richard Nixon and Kim Jong-un. Yet the striking feature was less the venue than the framing. The dominant question was how to implement AI, not whether its reach should be limited.
Robot anchors move from concept to stage
Xinhua and robotics company Ubtech introduced Xinyan and Xinyu, billed as the world’s first pair of “bionic humanoid robot news anchors.” A cinematic presentation showed them arriving in a flying saucer before delivering their first newscast. Their design involved Zhang Yimou, the film director who led the opening and closing ceremonies for Beijing’s 2008 Summer Olympics and 2022 Winter Olympics.
Yimou’s role included addressing the “uncanny valley”: the unease people can feel when a machine looks almost, but not quite, human. The anchors don’t fully pass as people at close range, but from the stage they appeared convincing enough to invite a practical question: could robots become newsroom colleagues?
That question matters beyond the spectacle. Newsrooms are under pressure to produce and distribute content across more formats, and AI tools are already changing that work. But the summit offered little discussion of journalism’s strained economics or of the costs and benefits of building frontier AI systems. The robot launch instead put the emphasis on capability and public presentation.
Two different routes to AI influence
The summit also reflected a widening split between the US and Chinese approaches to AI. The US strategy described in the discussion centres on frontier systems and restrictions on access to advanced chips and semiconductor technology. China is putting more weight on lower-cost models, open-source tools and related hardware, from humanoid robots to self-driving and flying cars.
The logic is to build a broad ecosystem of users and products that can compete with the reach of US technology companies. With strong support from the Chinese Communist Party, Chinese firms are producing models and devices intended to bring AI into many parts of daily and industrial life. President Xi Jinping, meanwhile, has called for global governance frameworks, as China seeks influence over the rules that shape AI beyond US jurisdiction.
These approaches carry different priorities. One centres on control of the most advanced systems; the other seeks wide adoption through accessible models and hardware. The summit’s tone suggested that China sees deployment as a strategic contest in its own right. That makes the pace of adoption a question of national direction as much as technical readiness.
There was little of the scepticism that often surfaces in Western conversations about whether AI should be embedded across everyday processes. In some Western debates, religious leaders and journalists have stressed the need to preserve human agency and scrutinise ethical and commercial incentives. At the Beijing gathering, those concerns did not take centre stage.
Disinformation tests the promise of control
The gap between ambition and caution becomes harder to ignore when AI-generated media enters the picture. Delegates discussed proposals such as digital watermarks to identify synthetic content, but attribution is only one part of the problem. As generation tools become cheaper and more capable, bad actors may find it easier to produce material that spreads before anyone can verify it.
A participant used a cybercrime example to underline how unpredictable the consequences of new technology can be. The comment referred to US$6.3 billion in cryptocurrency stolen from the Binance exchange, an incident attributed to TraderTraitor, a group linked to North Korea’s Lazarus Group. The comparison wasn’t a direct claim about AI-generated disinformation; it was a warning that technological change can enable outcomes well beyond the main contest between Washington and Beijing.
That is the practical challenge for media organisations. Detection tools may help, but they cannot by themselves restore trust or prevent manipulated content from circulating. Newsrooms will need to judge what they publish, explain how they verify it and adapt as the tools available to both journalists and malicious actors change.
Media influence extends beyond journalism
Debating journalism’s future in China carries an obvious contradiction. Independent journalism depends on freedom of expression, which is tightly restricted in China. In some countries represented at the summit, longtime leaders have also weakened the democratic checks that should hold power accountable.
Still, media is broader than journalism alone. It includes the technologies, production methods and distribution channels through which content reaches audiences. Those fields overlap with reporting, even though making content and doing journalism have different aims. China is developing technologies that will shape both, while its government is cultivating ties with media organisations around the world.
That influence deserves attention regardless of a person’s political outlook. Western leaders who attack the press may also weaken the relationship between journalism and the public, while China is working to expand its presence in global media. The summit made visible a country’s effort to shape not just AI products, but the conversations and institutions around them.
For delegates, seeing these arguments from another centre of power offered a useful counterpoint. AI’s costs and benefits will reach people unevenly, but no region will be untouched. The US-China competition is now playing out through models, chips, hardware and media systems, with consequences that could range from wider access to powerful tools to sharper risks for information quality.
Journalism sits at the intersection of those forces. It will be shaped by the technology and by the rules, business models and public expectations that determine how the technology is used. The piano-playing robots were a polished demonstration; the more consequential story is that many of the tools and platforms shaping journalism’s next chapter are being built in China.
===
Discussion
Sign in to join the discussion.