AI copyright deal: when tech giants write the rules of creativity

When algorithms become the new copyright police

Australia’s government is staring at a $50bn cheque—and the price tag is the soul of its creative economy. A leaked cabinet proposal would let AI firms scrape artists’ work in exchange for a $350m annual fund, framed as "compensation." The deal, if signed, won’t just set a precedent. It will codify what artists have feared for years: that copyright law is being rewritten in Silicon Valley boardrooms, not parliament.

The mechanics are simple. Tech giants get unfettered access to books, music, and art to train their models. In return, they dangle investment in datacentres—jobs, infrastructure, the kind of shiny GDP boost politicians crave. The $350m fund? A rounding error for companies posting quarterly profits larger than Australia’s annual arts budget. For creatives, it’s a Hobson’s choice: take the crumbs or watch your livelihood disappear into the black box of an algorithm.

This isn’t just an Australian problem. The UK’s Intellectual Property Office spent 2025 in closed-door talks with Google DeepMind and Anthropic, trying to square the same circle. The result? A "voluntary code" so vague it might as well be written in emoji. Meanwhile, British musicians and writers report their work appearing in AI outputs with no recourse—because the law hasn’t caught up, and the government is too busy courting Big Tech to care.

The asymmetry is staggering. When a human artist samples a song, they negotiate licenses, pay royalties, and face lawsuits if they step out of line. When an AI model ingests a million songs to generate a new track, the onus is on the original creator to prove harm. That’s not innovation. It’s a protection racket with better PR.

Japan’s robot army: when care becomes code

Japan’s new robotics strategy reads like a dystopian novel’s appendix. By 2040, 10 million robots will roam the country, some tasked with bathing the elderly, others flipping burgers in fast-food kitchens. The government calls it a "solution" to labour shortages. Critics call it a surrender of human dignity.

The plan, unveiled this week by Economy Minister Ryosei Akazawa, leans heavily on "physical AI"—robots that don’t just process data but interact with the physical world. A consortium of tech giants (SoftBank, Sony, Honda) will lead the charge, backed by public funds. The pitch? Robots will free humans from "3D jobs" (dirty, dangerous, demeaning). The reality? They’ll accelerate the erosion of care work’s already precarious wages.

Japan’s ageing population is a genuine crisis. But outsourcing compassion to machines is a choice, not an inevitability. The UK should pay attention. With NHS waiting lists stretching into 2028 and care homes chronically understaffed, the temptation to automate empathy will grow. Already, trials of "companion robots" in British care homes have shown mixed results—residents report less loneliness, but staff worry about the creep of "efficiency" metrics that treat human interaction as a cost to be minimised.

The deeper question is political. Japan’s robot army isn’t just about technology. It’s about who gets to decide what labour is worth. When a robot can bathe your grandmother for less than minimum wage, the pressure to cut human carers’ pay becomes irresistible. That’s not progress. It’s a transfer of wealth from the working poor to the shareholders of SoftBank.

Beavers vs. bureaucrats: when nature innovates faster than policy

While politicians haggle over AI and robots, Britain’s rivers are being quietly rewilded—by rodents. In the Wyre Forest, three beaver kits were born this week, descendants of a family reintroduced in 2024. In Ealing, a colony of beavers has stopped a train station from flooding. And in Shropshire, the River Kemp’s "wiggle" has been restored after a century of canalisation.

The results are undeniable. Beavers don’t just build dams; they create wetlands that filter water, store carbon, and provide habitat for endangered species. Their work is so effective that the UK government has quietly expanded reintroduction trials, despite opposition from landowners and farming lobbies.

Yet here’s the irony: while beavers get a free pass to innovate, human-led rewilding projects face years of planning permission battles. The difference? Beavers don’t ask for subsidies. They don’t lobby for tax breaks. They just get on with it.

The lesson for Britain’s innovation policy is stark. The most effective climate solutions aren’t always the ones with the glossiest PowerPoints. Sometimes, they’re the ones that have been around for 10 million years.

The quiet reckoning: what Britain can learn from Australia’s AI gamble

Australia’s $50bn AI deal isn’t just a cautionary tale. It’s a preview of Britain’s future.

The UK is already down this road. In 2025, the government fast-tracked planning permission for datacentres, waiving environmental reviews to lure tech giants. The pitch? Jobs, investment, a "global Britain" at the forefront of AI. The reality? A handful of high-paid tech jobs, a surge in energy demand, and a creative industry left to fight for scraps.

The difference is scale. Australia’s deal is brazen: a direct trade of copyright for cash. Britain’s approach is more insidious. A "voluntary code" here, a tax break there, a blind eye turned to copyright infringement. The effect is the same. Tech giants get to write the rules, and everyone else gets the bill.

The question for the new Labour government is whether it will break the cycle. So far, the signs aren’t promising. Andy Burnham’s state expansion plans include nationalising energy grids—but not the datacentres that are straining them. The party’s tech policy remains a love letter to Silicon Valley, with vague promises of "responsible innovation" and no mechanism to enforce it.

Australia’s artists are fighting back. Britain’s creatives are still waiting for their $350m. The clock is ticking.