Innovation Wars: Trump Guts NSF, China Plays Matchmaker

Innovation policy fractures across three continents this week as Trump fires the NSF board, Beijing pairs 680,000 researchers with firms, Canberra taxes Big Tech.

Innovation Wars: Trump Guts NSF, China Plays Matchmaker
Photo by Ricardo Gomez Angel on Unsplash

TL;DR

Three governments moved on innovation policy in the same news cycle, and they could not have moved in more different directions. Washington is dismantling its scientific guardrails. Beijing is herding 680,000 researchers into corporate arms. Canberra is forcing Silicon Valley to subsidise the journalism it scrapes. The UK, neither imperial laboratory nor command economy, is left watching three rival operating systems for innovation try to outrun each other.

Why is Trump firing the National Science Foundation board?

The Guardian reports that members of the National Science Board — the independent body that oversees the NSF — received an email on Friday from the Presidential Personnel Office, sent "on behalf of President Donald J Trump," informing them their position was "terminated, effective immediately." Critics quoted by the paper called the move "a dangerous attack" on US innovation.

The NSF is not a partisan trinket. It funds the basic research — the unfashionable, decades-long, no-product-at-the-end kind — that produced the internet, GPS, the polymerase chain reaction, and a long catalogue of technologies that American firms later monetised. Sacking its independent overseers does not end that work overnight. It ends the firewall between political appointees and the grant pipeline.

For British research leaders, the implication is double-edged. A weakened, politicised NSF is bad news for the global scientific commons that UK labs depend on for collaboration and data. It is also, bluntly, an opportunity. American post-docs already lobbied by European recruiters now have a fresh argument for crossing the Atlantic. UKRI has not, at the time of writing, said anything publicly. It probably should.

How does Beijing's 680,000-researcher matchmaking scheme work?

While Washington kicks out its science board, Beijing is plugging its scientists directly into the production line. Nature reports that China's latest commercialisation push aims to match 680,000 innovators with companies — a centrally orchestrated handshake between universities, research institutes and industry. The journal contrasts the approach explicitly with the United States, "where market forces drive innovation."

That framing is a little generous to the US these days, but the structural point holds. China is treating innovation as logistics: identify the researchers, identify the firms, build the matching engine, count the patents and products at the end. It is the same playbook that delivered electric vehicles and solar panels at world-shaping scale. Whether it works as well for frontier basic science — the kind that rewards heretics over planners — is the open question.

For the UK, the contrast cuts close. Domestic policy oscillates between pleading with universities to "spin out more" and lamenting that they don't. Beijing's answer is to make the introductions in bulk. London's answer, so far, is the Advanced Research and Invention Agency and a lot of speeches. The gap in sheer industrial seriousness is widening.

Why is Australia making tech giants pay for news?

The Guardian reports that the Albanese government has proposed a 2.25% levy on the local revenues of digital platforms that fail to strike commercial deals with Australian publishers. Anthropic — sorry, Albanese — argued that tech giants "should not be able to exploit the work of journalists to boost profits."

Google's response, per the Guardian, was to reject the case for reform and complain that AI platforms were excluded from the scope. Meta, which runs Facebook and Instagram, called the government's position "simply wrong."

The complaint about the AI carve-out is the more interesting one. Google's broader business model is increasingly an answer machine, not a referral machine — one that ingests journalism and returns it pre-digested to users who never click. Canberra's draft levy targets the old extraction model. The new one, in which large language models train on and resell newsroom output, walks free.

Britain has its own Digital Markets, Competition and Consumers regime and its own restless publishers. The Australian template — pay or be levied, with no opt-out — will be studied carefully by ministers. The signal to Mountain View is unsubtle: regulators are no longer asking nicely.

The pressure is also coming from inside the building

Two further data points round out the picture. The Register reports that "AI deflation" is now showing up in the revenue lines of India's big four IT services firms — software work being done faster and cheaper as models absorb more of the routine coding load. And Japan Airlines, per the Guardian, will trial humanoid baggage handlers at Haneda from May, on the back of a chronic labour shortage.

Stitch them together and the storyline tightens. Innovation is not a neutral process happening to passive economies. It is being actively steered — by White House diktat, by Chinese central planning, by Australian tax law — and it is hollowing out service-sector business models in real time.

What to retain

Three policy moves, three philosophies. Washington kneecaps the institution that made it scientifically dominant. Beijing industrialises the introduction between scientist and CEO. Canberra hands tech giants the bill for the journalism they consume. None of these is the British model. None of them yet has to be. But pretending the old transatlantic consensus on how innovation works is still operating would be the most dangerous oversight of all.