China’s AI ‘Digital Iron Curtain’: Trade Secrets Law or National Security Lockdown?
China’s new June 1 trade-secret rules protect algorithms, source code and training data, showing that Beijing now treats AI as strategic infrastructure.
China’s new trade-secret rules mark a turning point in the global AI race. As of June 1, Beijing’s updated framework strengthens protection for algorithms, source code, training datasets and other non-public technical assets. Viral posts are calling it a “digital iron curtain.” That phrase is dramatic, but the direction is real: China is treating AI knowledge as strategic property that must not leak.
The rules modernize China’s administrative regime for protecting trade secrets. In practice, this means companies have stronger legal tools against hacking, unauthorized scraping, employee theft, disclosure of model weights, stolen datasets or the transfer of sensitive technical know-how. Fines and enforcement tools are being strengthened. Employers and state regulators now have a clearer basis to treat core AI systems as protected assets.
This is not happening in isolation. The United States restricts advanced chips. China restricts technology leakage. Washington blocks hardware. Beijing locks down people, code, datasets and algorithms. Both sides are treating AI less like a normal business sector and more like a weapons-adjacent strategic domain.
The comparison to a weapons program is uncomfortable but increasingly accurate. AI models can support military targeting, cyber operations, drone autonomy, intelligence analysis, propaganda, surveillance, logistics and industrial automation. The data used to train them and the code used to deploy them are not just commercial assets. They are national power inputs.
For Chinese companies, the new rules may be welcome. Firms like Alibaba, DeepSeek, Huawei, Baidu and others have strong incentives to protect proprietary systems from theft, former employees, foreign rivals and state-linked espionage. Stronger trade-secret protection can encourage investment because companies know their core assets are legally defendable.
But there is another side. If AI assets become heavily securitized, research collaboration may suffer. Engineers may face travel limits or legal risks. International partnerships may become harder. Academic openness may shrink. Foreign investors may worry that data, models or technical cooperation are trapped behind regulatory walls. The boundary between protecting innovation and restricting knowledge can blur quickly.
This is why the “digital iron curtain” phrase resonates. It suggests the world is moving toward two AI blocs: one shaped by U.S. chips, cloud platforms and export controls; another shaped by Chinese data rules, domestic compute and state-backed industrial policy. The internet once promised borderless information. AI may be returning technology to borders, passports and national security reviews.
The question is whether this helps or hurts China. Protection may prevent theft and retain strategic advantage. But too much control can slow creativity. AI progress depends on talent movement, open research, benchmarking and global communities. A system that locks everything down may become secure but less dynamic. A system that leaves everything open may become innovative but vulnerable.
The headline says China dropped a digital iron curtain around AI. The more precise conclusion is that Beijing has decided AI secrets are no longer merely corporate property. They are national assets. In the U.S.-China race, the next battle may not only be over chips. It may be over who controls the code, data and people behind them.