How the U.S.-China Rivalry Is Putting the Internet at Risk

Video calls, streaming services, social media: In our wireless age, they all seem to simply float in the cloud. It’s easy to forget they are transmitted along actual physical wires running across the seabeds of our planet. Spanning nearly 750,000 miles, these 400-plus subsea cables handle more than 95% of internet traffic, whether it’s the news, SWIFT transactions, or government communications. Cables are the “out-of-sight arteries of globalization,” as Surabhi Ranganathan, a law professor at Cambridge University and expert on oceans, puts it. But the U.S.-China cable race is putting this essential infrastructure at risk, as existing institutions tasked with its protection seem unprepared to step in.

A new Asia-Europe data link is the most recent example. In February, U.S. company SubCom landed the subsea cable project for $600 million, only three years after HMN Tech, a formerly Huawei-owned Chinese company, nearly won the contract. A global consortium of telecom firms, including China Telecom and Microsoft, went for SubCom instead, even though the Chinese bid was substantially cheaper. The move came after fears of espionage had led the U.S. government to apply significant pressure on telecom firms not to back HMN Tech, reportedly under threat of sanctions. Yet as transpired in April, HMN Tech is now simply developing its own $500-million Asia-Europe cable passing via Singapore and the Middle East to France—in each of which the Chinese-led consortium has business agreements with national telecom providers.

This is not an isolated instance. The U.S. Department of Justice’s Team Telecom regularly interferes to keep Chinese companies from winning bids or laying direct U.S.-China cable links. China in turn allegedly cut off subsea internet cables at Taiwan’s Matsu island, which is closer to the mainland, in early February. Some have suggested it was a test case for China to try similar moves elsewhere. Could geopolitical cable competition, as security analysts suggest, fragment the ostensibly one-world internet infrastructure? Are we nearing a “Cold War under the sea”?

It’s not that simple. Data flows have never been truly “global.”

The first deep-sea cable connections were laid in the 1850s, and by 1866 the SS Great Eastern had laid the first durable transatlantic cable. Instead of bringing world peace or a “Victorian internet,” telegraphs connected imperial outposts more than anything. In subsequent decades, a boom in deep-sea cables greased the engines of European imperialism. Contemporaries referred to them as “the nervous system of the British Empire.” Gutta-percha, the rubber used to insulate subsea cables, became a prized commodity extracted from latex-producing colonies like Singapore and Borneo.

More From TIME

To smooth out technical complications, government delegates created the International Telegraph Union (ITU) in 1865, which became one of the world’s first international organizations. ITU would go on to adopt universal standards for Morse code, tariffs, and accounting rules. Other steps were taken to protect deep-sea cables; in 1884, delegates of 30 governments gathered in Paris and adopted the Convention for the Protection of Submarine Telegraph Cables. Yet these regulatory efforts failed to depoliticize telecommunications. Telegraph lines remained fiercely contested, and governments were unable to agree on protections against wartime sabotage.

The 1982 U.N. Convention on the Law of the Seas (UNCLOS) complemented the ITU by protecting companies’ “freedom to lay submarine cables and pipelines” in international waters. Yet the U.S. never signed the Convention, and both the U.S. and China have their own provisions that contradict UNCLOS. For example, UNCLOS protects the right of companies to lay and maintain cables even in a state’s exclusive economic zone (EEZ) unless this interferes with that state’s other rights. Telecom representatives are thus eager that the U.S. ratify the Convention, so that they can operate off America’s shores with less geopolitically-motivated interference.

Fast-forward to today, the ITU, which is now a U.N. agency, is still meant to oversee globalization’s electric arteries. While the ITU sets technical standards for data cables, key decisions are still taken behind closed doors in national capitals and in corporate fora.

The disparate number of players—from governments to private businesses to industry forums—complicates the ITU’s coordination and agenda-setting powers. As does the messy and unclear ownership structures of multi-company consortia that typically build, own, and operate subsea cables, as well as jurisdictional questions given that they span both domestic and international waters.

Still, ITU and UNCLOS provisions that facilitate laying cables in the high seas in the absence of a robust regulatory environment are encouraging the kind of tit-for-tat disputes over the world’s subsea cables seen earlier this year. This not only puts actual internet access at risk but also fuels conflict in geopolitically-strategic locations such as the South China Sea.

In March this year, the U.S. Congress passed the UnderSea Cable Control Act, pitting “American superiority in undersea cable capabilities” against “China’s economic and military reach.” Such language is an unnecessary provocation and another reminder that we’re far way from a global data network based on cooperation and independent regulation.

The ITU’s Strategic Plan for 2024-2027 calls for equal internet access for everyone. In a world where a vast array of services and communications depend on it, that is a laudable goal. But the actions of cable operators and governments suggest that the infrastructure underpinning the world’s data networks may actually be going in the opposite direction, in ways that could ultimately jeopardize this goal.

The ITU has to adapt and may well need to assume a stronger role. This would include not shying away from imposing clear rules about subsea cables, and actively reining in companies or their state backers. Whether it can do so remains the big question.