China's tech export strategy: EVs, open-source AI and the fight over standards

In June 2026, a purpose-built LNG-powered bulk carrier docked at the Port of Melbourne on its first voyage to Australia. On board the BYD Zhengzhou were 5,000 newly built electric vehicles, the largest single EV shipment ever sent to Australia, according to RNZ's report on the arrival.

The shipment is a useful marker. Australia has no domestic car industry and few trade barriers, so RNZ's analysis describes it as a litmus test for global consumer taste. What it shows is a market tilting fast towards Chinese brands.

Sales figures tell the story

In the first half of 2026, according to the same RNZ analysis, sales for every major Japanese manufacturer bar Honda fell in Australia. Mazda shed 17 percent year on year, Toyota 21 percent, Subaru and Mitsubishi 25 percent each, and Nissan 32 percent. Chinese suppliers moved the other way: BYD jumped 124 percent, Chery 77 percent and Geely 495 percent. All three started from much smaller bases, which exaggerates the percentage gains.

Last month, almost a quarter of all cars sold in Australia were fully electric, up from 7 percent a year earlier. Add hybrids, and EVs now match petrol and diesel cars. Riz Akhtar, founder of the research group carloop, told RNZ he drives a Chinese EV that costs around A$40,000, and that an equivalent Japanese or European model would cost more than 25 percent more. "Consumers aren't stupid. They're not going to spend money they don't need to," he said.

The pressure is visible elsewhere. Toyota vice-chairman Koji Sato told an annual supplier meeting in March that Japanese producers needed to cooperate on costs. "Unless things change, we will not survive," he said, according to RNZ. Volkswagen chief Oliver Blume has foreshadowed lay-offs of up to 100,000 workers, and VW's share price has halved since his predecessor was removed four years ago, the report adds.

Open source as an export channel

The second front is software. Writing in The Conversation, a China expert and open-source researcher argues that Beijing has folded open-source development into its national technology strategy, and that the countries which set the technical and legal standards will hold an edge for decades.

The history runs from Red Flag Linux, built by a government-run software institute as an alternative to Microsoft Windows, to Kylin, an operating system created by a Chinese military university for the People's Liberation Army. After Washington blacklisted Huawei, Google revoked its access to licensed services. Huawei used Android's open-source code to keep its phones running, then built HarmonyOS NEXT and an open-source version called OpenHarmony.

Today the Ministry of Industry and Information Technology backs domestic platforms including the state-guided OpenAtom Foundation and Gitee, a Chinese alternative to GitHub. Domestic systems such as openKylin and openEuler are described as securing China's IT infrastructure, while work on the open RISC-V chip standard reduces dependence on foreign semiconductor intellectual property.

Chinese telecommunications companies, already leaders in 5G, are working with an international alliance to shape 6G, the next-generation mobile standard.

Licensing is where the strategy gets contentious. Major Chinese AI models including Qwen, DeepSeek, Kimi and GLM are generally open-weight rather than fully open-source: the software and final parameters are published, but not the training data or methodology. Some licences carry extra conditions. Moonshot's Kimi K3 licence requires large licensees to display the model name prominently, and MiniMax M3's community licence lists uses prohibited by "applicable laws or regulations", leaving open which jurisdictions apply.

Bilingual licences developed under Chinese government guidance, the Mulan Permissive Software License version 2.0 and the OpenAtom Model License version 1.0, give Chinese users more access and Chinese courts more control. Both are written in English and Chinese, but the Chinese text prevails in conflicting interpretations, especially in Chinese jurisdictions.

The White House argument

Those models have landed in Washington like a rock. MIT Technology Review reported on 20 July that David Sacks, Trump's AI and crypto czar until March, called Anthropic's models "lobotomized" and "woke", while Pentagon official Emil Michael called OpenAI's new head of strategic futures a "supreme village idiot". The trigger was Kimi, a free model from Moonshot that appeared to rival paid models from OpenAI and Anthropic.

Sacks argued that Chinese models are popular because they come with fewer restrictions, setting aside their built-in state censorship, and criticised US AI companies that "want the government to eliminate their open source competition". Dean Ball, a former Trump AI adviser now at OpenAI, called the new White House review process a "de facto licensing regime for frontier AI". Michael replied that the government would work through "the democratic process not some Deep State scheme".

The Guardian reported on 1 August that the split runs through industry too. Chip makers and companies worried about OpenAI and Anthropic's dominance favour open models; Anthropic and OpenAI, facing profit pressure, argue Chinese models pose security risks. Microsoft, Nvidia, Palantir and Meta signed a letter urging lawmakers not to restrict open models, and Nvidia chief executive Jensen Huang went to Capitol Hill to lobby. Treasury secretary Scott Bessent suggested sanctioning Chinese AI firms over alleged intellectual property theft, while commerce secretary Howard Lutnick received letters from startup founders asking him not to cut off access.

Trump, asked about safety restrictions, said: "We have to be careful in both ways. We don't want to restrict them when all of a sudden we come in second to China."

On the same day, the Federal Communications Commission banned humanoid robots from China, alleging they could steal data, surveil US citizens and threaten manufacturing supply chains. A report from The Information that China had begun mass production of specialty AI chips erased $1tn in market value from other chip makers, The Guardian said.

Batteries, and the cost of cutting ties

The third front is hardware, and here the policy is blunter. In late August 2026, the Trump administration declared a national emergency that effectively bans Chinese batteries from grid-scale energy storage, MIT Technology Review reported on 10 September. The order bans installation of "any foreign-produced bulk-power system electric equipment" posing a national security risk and names battery storage systems, inverters and transformers.

Analysts quoted by the publication expect delays. Shan Tomouk of Benchmark Mineral Intelligence called an outright ban "a bit of a surprise" that "does create a bit of concern for domestic players". BloombergNEF analysis says projects may stall while developers await Department of Energy guidance due by the end of the year, and Isshu Kikuma said that in the worst case projects could be cancelled. Enforced to the letter, the order would mean removing most installed battery storage from the US grid.

Domestic capacity might meet demand around 2030, though some factories may not ramp up fully, pushing real supply past that date. Tariffs on batteries rose to 25 percent in January from 7.5 percent, and from 2026, 55 percent of the material cost of new storage projects must come from outside China and other restricted countries to qualify for tax credits.

None of this happens in isolation. Huawei opened its US trial in Brooklyn federal court on Wednesday, where prosecutor Taylor Stout told jurors: "Theft, lies, cover-up." The government alleges Huawei stole trade secrets from five US companies, including router source code from Cisco and a robotic arm from T-Mobile. Defence lawyer Brian Heberlig countered: "It's about competition, not conspiracy. Innovation, not theft."

China's foreign ministry said it "firmly opposes the US side's suppression and containment of Chinese enterprises". The trial is expected to last three months.