China throws money after tech, while Trump pivots away

Politico

12-03-2025

With help from Alfred Ng
As the world's two biggest economies race to dominate technologies of the future, like artificial intelligence and quantum computing, they're betting on opposing industrial strategies.
China just previewed a new state VC fund, intended to support startups focusing on cutting-edge fields like semiconductors, quantum, AI and renewable energy, with an eye-popping target of $1 trillion yuan, or roughly $138 billion.
President Donald Trump, meanwhile, used his first address before Congress to attack America's largest attempt at directly investing in high-tech manufacturing. He called for lawmakers to undo the CHIPS and Science Act, the 2022 law that has been pouring some $50 billion into rescuing the domestic semiconductor sector.
'China is dialing its industrial policy up to 11 while the U.S. is rolling back its industrial policy,' is how Gerard DiPippo, the acting associate director of the RAND China Research Center, put it.
The conventional wisdom lately has been a little different. For a while, it looked like America was re-embracing high-tech government intervention. The CHIPS Act and the Inflation Reduction Act, another signature Biden-era policy that Trump wants to repeal, represented a return to industrial policy on a scale the U.S. had not seen in decades. Trump himself ordered the creation of the U.S.'s first sovereign wealth fund, an ill-understood idea that he's said could invest in state-of-the-art manufacturing hubs and may be closer to what China has in mind.
But in the broader sense, U.S. tech policy largely amounts to assuming the private sector will carry the ball. More structured, long-term government investments have been subject to powerful political winds that can change from administration to administration, and recently even month to month. Trump's upcoming budget request is expected to make major cuts to the science agencies that conduct basic tech research. And his U.S. sovereign wealth fund — first suggested by former President Joe Biden last fall — is still a purely 'look into it' proposal.
Meanwhile, China plows forward with a strong emphasis on industrial policy — one that, on tech, actually dates over a full decade. Its ongoing 2015 'Made in China 2025' plan was designed to upgrade its manufacturing of high-tech goods and achieve self-sufficiency in the most critical supply chains.
The dueling statements last week were a striking reminder of the persistency gap between Beijing's continuous, state-backed policies versus the U.S.'s more fragmented and sometimes reactive market-driven approach, said Kyle Chan, a researcher at Princeton University who writes the popular newsletter 'High Capacity' on industrial policy.
Consistency matters a lot in certain industries like semiconductors, where factories cost so much that some chipmakers bet their entire future when they plan a new investment. China's semiconductor-specific war chest, known as the 'Big Fund,' goes back to 2014 and started spending its third installment in January.
There's no true U.S. counterpart to Chinese industrial policy. 'The closest equivalent would be maybe DARPA or In-Q-Tel, which supports the intelligence community, but those are dealing with much smaller investments,' said Will Rinehart, an economist specializing in tech policy at the American Enterprise Institute. 'Just generally speaking, the U.S. doesn't have equivalently sized, government-owned investment funds.'
At times, it has seemed to inch closer in recent years. The Obama Department of Energy extended large loans to startups like Abound Solar and the controversial Solyndra to help them bring advanced green technologies to market. Biden's ability to get CHIPS passed with bipartisan votes marked a notable shift among Republicans, a party that traditionally balked at industrial policy as inefficient government interference.
Will Trump's reversion make a difference? That hinges on broader questions, like whether industrial policy even works — and what it means for two vastly different countries to make it work.
Compared to other advanced economies, the U.S. has historically been the most averse to any consistent use of industrial policy. 'The U.S. is a market-oriented, commercially-oriented system. China is increasingly actually a national security and capacity-oriented system,' said DiPippo. 'We're playing different games, and that means the rules are different, and it's hard to say who wins if you're both playing by different rules.'
The effectiveness of industrial policy depends, too, on the technology in question. For AI, the U.S.'s leadership in the best models and chips largely came from the private sector. Even China's notable win — DeepSeek, the powerful-and-cheap new AI model — is no obvious poster child for industrial policy. 'There might have been some sort of broader support in terms of data access,' said DiPippo. 'But it seems like they were just a scrappy firm that had some really good engineers.'
The difference could matter more in industries like chip manufacturing and quantum computing, where national security implications may require long-term investments to arrive faster than the market demands. That's where some Republicans continue seeing a role for subsidies — as recent hesitancy toward Trump's demands to repeal Biden-era laws shows.
Trump's approach hasn't exactly been a model of free-market thinking, of course: Like Biden, he has used federal power as a tool to drive economic outcomes. Unlike Biden, his favorite lever is tariffs, which he treats as a blunt force to get desired policy outcomes out of chipmakers like TSMC, 'for free' as he says.
While it may be where the comparison ends, that tactic to prod the private sector into action is one similarity the two countries still appear to share.
'There's a very strong, almost like parallel between what Trump is trying to do and how he's trying to do it, and honestly, what China's whole sort of industrial policy approach has been,' Chan said. 'Use your greatest asset — which is access to your market — to bring foreign companies, to bring your own companies and keep them within your own borders.'
TAPPING THE BRAKES
The California Privacy Protection Agency fined Honda $632,500 in a settlement resolving alleged unfair data collection practices, as my colleague Tyler Katzenberger reported from Sacramento this morning.
The agency claimed Honda made it difficult for customers to opt out of having their data sold or shared with advertisers in violation of state law. Along with the fine, Honda agreed to make it easier for owners to opt out, and is required to consult a designer to review its methods for privacy requests.
The settlement marks the first time California went after an automaker, but the industry has found itself under the microscope as of late over how it handles people's data. Texas and Arkansas sued General Motors for alleged privacy violations, while the Federal Trade Commission banned the company from sharing drivers' data to consumer reporting agencies for five years.
Cars over the last decade have quietly become a major source of data for companies to collect and sell, siphoning information on a car's location, a person's phone contacts, and even the air pressure in the tires. Digital Future Daily took a look at this trend and the potential data risks back in 2022.
CRYPTO RULE HEADED FOR REPEAL
A crypto regulation requiring financial transparency may soon be revoked after the House voted to repeal it Tuesday.
The Biden-era rule required certain cryptocurrency brokers to report digital asset transactions to the Internal Revenue Service and collect customer information, which supporters said would provide more visibility into unreported transactions, POLITICO's Bernie Becker reported.
The House voted 292-132 to repeal the rule, and it is headed to the Senate for final action before it goes to President Trump's desk, who is expected to sign it.
POST OF THE DAY
THE FUTURE IN 5 LINKS
Stay in touch with the whole team: Derek Robertson ([e-mail]); Mohar Chatterjee ([e-mail]); Steve Heuser ([e-mail]); Nate Robson ([e-mail]); Daniella Cheslow ([e-mail]); and Christine Mui ([e-mail]).