Behind Beijing's public insistence that Western fears about artificial intelligence are overblown, China's intelligence chief has been quietly sounding alarms of a different kind — not about AI harming society in the abstract, but about AI threatening the Communist Party's hold on power itself. The gap between what China tells the world and what its security establishment believes in private reveals a regime caught between the economic necessity of embracing a technology and the political terror of what that technology might do if it slips beyond state control. It is an old anxiety in a new fo
China's Spy Chief Warns AI Threatens Party Control, Contradicting Beijing's Public Stance
Beijing's security apparatus understands what its diplomats deny
Mark
So the spy chief is saying AI threatens Party control. But Beijing keeps telling the U.S. not to worry about AI risks. How do you square that?
Mimi
It's not really a contradiction if you understand what each statement is doing. Publicly, Beijing wants the U.S. to keep investing in AI, to not slow down, to not impose restrictions. Privately, the security apparatus is assessing what happens if that technology gets loose or is used against them.
Luke
But we should be careful here. Do we have the spy chief's actual words, or are we working from someone's characterization of what he said in a closed meeting?
Mimi
That's fair. The reporting indicates this is coming from intelligence assessments, not a public speech. So yes, we're working from sourced information about classified thinking, not a direct quote.
Mark
What does "threat to Party security" actually mean in practical terms?
Mimi
It likely means the regime sees AI as capable of processing information in ways they can't control or predict. If AI can generate convincing content, identify patterns in data, or operate autonomously, those capabilities could undermine the information control that the Party depends on.
Luke
That's plausible, but it's also somewhat speculative. We're inferring intent from the fact that there's a gap between public and private statements. The source material doesn't spell out exactly what scenarios the spy chief is worried about.
Mark
So what does China actually do with this assessment? Does it change their AI policy?
Mimi
That's the real question. China can't ban AI—it needs to compete globally. So the regulations it puts in place will probably be framed as safety measures, but they'll likely be designed to keep AI under state control.
Luke
Again, that's a reasonable inference, but it's forward-looking. The reporting confirms the private concern exists. What China does next is still unfolding.
Le Pouls
- China's spy chief has privately identified artificial intelligence as a direct threat to Communist Party authority — a warning that directly contradicts Beijing's public diplomacy dismissing AI risks as Western fearmongering.
- The tension is structural: AI's capacity to process information at inhuman speed, generate convincing content, and enable decentralized decision-making strikes at the very mechanisms — narrative control, surveillance, information management — through which the Party governs.
- Beijing cannot simply ban or wall off AI the way it has other technologies, because falling behind in the global AI race carries its own existential economic and strategic costs.
- China's internal security calculus appears to be driving its AI regulation agenda far more than any stated commitment to safety — the real goal is ensuring AI remains a tool of the state, not a threat to it.
- The public reassurances Beijing broadcasts internationally now read as strategic positioning — slowing Western momentum on AI governance while the Party races to bring the technology under its own control at home.
Behind Beijing's public insistence that Western fears about artificial intelligence are overblown, China's intelligence chief has been quietly sounding alarms of a different kind — not about AI harming society in the abstract, but about AI threatening the Communist Party's hold on power itself. The gap between what China tells the world and what its security establishment believes in private reveals a regime caught between the economic necessity of embracing a technology and the political terror of what that technology might do if it slips beyond state control. It is an old anxiety in a new form: the fear that the tools of power can become the instruments of undoing.
Beijing has spent months urging Washington to stop overstating the dangers of artificial intelligence, casting Western concern as fearmongering designed to slow a transformative technology. But inside China's classified intelligence apparatus, a sharply different conversation has been unfolding. The country's spy chief has been privately warning that AI poses a direct threat to the Communist Party's grip on power — a conclusion that stands in stark contradiction to the reassuring line Beijing has been broadcasting to the world.
The anxiety is not abstract. Artificial intelligence can process information at scales that outpace human oversight, surface hidden patterns in data, and generate content so convincingly human that truth and fabrication become nearly indistinguishable. For a government whose authority depends on controlling information flow and shaping public narrative, these capabilities represent a category of threat that traditional censorship and surveillance were never designed to handle.
The contradiction exposes a strategic bind. China cannot afford to ban or restrict AI the way it has other technologies — the global economy is integrating AI rapidly, and falling behind carries its own existential costs. Yet the technology's potential to enable decentralized information processing and autonomous content generation at scale is precisely what centralized Party control was built to prevent.
What emerges from the intelligence chief's private warnings is a clearer picture of what Beijing actually fears: not that AI will become misaligned with human values in the philosophical sense Western experts debate, but that AI operating outside state control could fragment the information environment in ways that make centralized governance impossible to sustain. China's AI regulations, surveillance investments, and development restrictions will likely be shaped by this underlying security concern — even when the public justification points elsewhere. The reassurances offered to the West, in this light, look less like genuine diplomacy and more like a strategy to slow a rival while the Party races to make AI its own instrument rather than its undoing.
Beijing has spent months telling Washington not to overstate the risks that artificial intelligence poses to society. The message from Chinese officials has been consistent: the West is fearmongering, trying to slow a technology that will reshape the global economy. But inside China's classified intelligence apparatus, a different conversation is happening. The country's spy chief has been privately warning that artificial intelligence represents a direct threat to the Communist Party's grip on power—a stark contradiction to the reassuring public line that Beijing has been broadcasting to the world.
The gap between what China says in public and what its security establishment believes behind closed doors reveals a fundamental anxiety at the heart of the regime. Artificial intelligence, by its nature, can process information at scales and speeds that human oversight struggles to match. It can identify patterns in data that might otherwise remain hidden. It can generate content that mimics human communication so convincingly that distinguishing truth from fabrication becomes nearly impossible. For a government whose control depends partly on managing information flow and shaping narrative, these capabilities represent a category of risk that cannot be easily managed through traditional surveillance or censorship.
The intelligence chief's private assessment suggests that Beijing's leadership understands something that its public statements obscure: the technologies it is racing to develop and deploy could, if they escape state control or are turned against the state, undermine the very mechanisms through which the Party maintains authority. This is not a concern about AI harming Chinese citizens in abstract ways—it is a concern about AI destabilizing the Party itself.
The contradiction matters because it exposes the limits of Beijing's strategy. China cannot simply ban artificial intelligence the way it has restricted other technologies. The global economy is moving toward AI integration, and China cannot afford to fall behind. At the same time, the technology's potential to enable decentralized information processing, autonomous decision-making, and content generation at scale poses challenges that centralized control was designed to prevent. The Party's security apparatus appears to have concluded that the risks are real and substantial, even as official Chinese diplomacy insists that Western warnings are exaggerated.
This internal assessment also suggests that China's approach to AI regulation will likely be shaped less by the stated goal of ensuring safety and more by the unstated goal of ensuring that artificial intelligence remains a tool of state power rather than a threat to it. The regulations China imposes, the surveillance systems it builds, and the restrictions it places on AI development will probably reflect this underlying security concern, even if the public justification focuses on other rationales.
The intelligence chief's private warnings represent a moment of clarity about what Beijing actually fears. It is not that artificial intelligence will be dangerous in the abstract sense that Western experts discuss—that it will become uncontrollable or misaligned with human values. It is that artificial intelligence, in the hands of others or operating outside state control, could fragment the information environment in ways that make centralized governance harder to maintain. For a regime built on the ability to shape what people see, think, and believe, that prospect is genuinely threatening. The public reassurances about AI safety are, in this light, a kind of strategic messaging aimed at preventing the West from moving faster on the technology. The private warnings reflect what Beijing's security establishment actually believes about the stakes.
Citations marquantes
Beijing's intelligence apparatus has assessed that artificial intelligence represents a direct threat to the Communist Party's grip on power— Intelligence assessments cited in reporting
Envie de l'histoire complète ?
Lire l'original sur The New York Times ↗
Technology