House of Lords Votes in Favor of AI Bill Amendment, Compelling Businesses to Disclose Copyrighted Work Utilized in Artificial Intelligence Model Training
In a sentence that wouldn't be out of place in a half-baked sci-fi novel, the Lords are standing up against the UK government's questionable plan on AI training. This ain't the sleek future I envisioned, 'bout flying electric cars and all that jazz.
According to the Guardian's gossip, the UK government is jonesing to educate AI on copyrighted materials, as long as the owners don't protest. Critics are crying foul, pointing out that this opt-out strategy is more of a headache than it's worth. You might not even know if your work is being snatched up by an AI, let alone have the smarts to dispute it.
The government's clumsy defense is that we're putting a halt on creative and technological growth by enslaving these AI beasts with restrictions. They're likely echoing other countries' worries over China's supremacy in DeepSeek LLM and similar AI leaps. Still, it smells more like free labor than technological advancement to us artists.
While the government's eyes are on global progress, artists like Paul McCartney, Jeanette Winterson, Dua Lipa, and the Royal Shakespeare Company are worried about their work being pilfered. They've even penned an open letter to the government, begging them not to give their labor away to a handful of zombie tech companies. A proposed amendment by crossbench peer Beeban Kidron aimed at requiring AI firms to spill the beans about their copyrighted material usage was passed with a resounding "hell yeah" from the Lords, but met with stubborn opposition from the government.
Lady Kidron gave the government a piece of her mind: "So let's reject the idea that we're against technology. Creators aren't denying the creative and economic value of AI, but they're crying foul when asked to build it for free with their work, only to rent it back from thieves." She continued, "This is an attack on the British economy, happening on a massive scale in a sector worth £120bn to the UK. It's a vital part of our industrial strategy and of immense cultural importance."
Assuming the government doesn't listen to its artists and doesn't give artists a fair cut, at least this battle has a silver lining: it'll be easier for people to discover if their work has been ripped off by AI. Plus, the mere acknowledgment of these training methods is a tiny credit to creators—if these firms are honest about their reporting and don't find some other legal loophole first.
The Biggest Gaming News, Reviews, and Hardware Deals
Stay on top of the latest gaming news, reviews, and hardware bargains, as picked by the PC Gamer team.
*Best Gaming Monitor: Crisp, pixel-perfect panels.Best High Refresh Rate Monitor: Screaming-fast refresh rates.Best 4K Monitor for Gaming: High-resolution, only 4K.Best 4K TV for Gaming*: Gaming-ready, big-screen 4K TVs.
Hope CorriganWith a decade of talking about games under her belt, Hope started off at the Aussie Nintendo fan site Vooks.net. Since then, she's blabbed about games and tech for various publications, such as Techlife, Byteside, IGN, and GameSpot. You can also find her breathlessly discussing her own creations right here at PC Gamer. When she's not crafting the creations of others, Hope is creating her own fantastical, ambient sci-fi sounds for a future podcast. Check it out here.
Enrichment Data:
Overall:
The current standoff in the UK's legislative process involves a significant amendment passed by the House of Lords—namely Amendment 49B — which required greater transparency from AI developers regarding their use of copyrighted content in training their models. The House of Commons, however, rejected this amendment, raising concerns about potential increased public spending and the broader impact on the bill.
Vote and Amendment Details:
- On May 13, 2025, the House of Lords passed Amendment 49B by a vote of 272 to 125, which requires:
- AI developers to list categories of content (text, audio, images, video) used in training and fine-tuning.
- Confirmation to rights holders, upon request, whether their individual works were included in the training datasets.
- Identification of automated tools, such as web crawlers or scrapers, used to collect data and clarification about their purposes.
- The vote also included a safeguard for small and micro-entities under UK law.
Outcome in the House of Commons:
- Despite the Lords' approval, the House of Commons rejected the disclosure requirement, citing "financial privilege" because the amendment could increase public spending without Commons' approval.
Overall Status:
The bill, including the Lords' amendment on AI transparency, remains in legislative limbo. The bill will return to the House of Commons for further deliberation, and if the Commons removes the amendment, it may lead to another confrontation with the House of Lords. Critics argue the Lords' amendments could stifle AI innovation by imposing technically difficult and costly legal requirements, potentially derailing the bill's passage, which is crucial for the UK's AI ecosystem.
- The Lords' amendment, passed by a vote in the House of Lords, requires AI developers to disclose the categories of content used in training their models.
- The proposed law seeks to enable rights holders to confirm if their individual works were included in the training datasets of AI models, which has raised concerns among politicians and artists.
- Despite the attempt to bring transparency in AI training policies, the House of Commons has rejected the disclosure requirement due to concerns over additional public spending.
- The standoff between the House of Lords and the House of Commons could potentially derail the bill's passage, impacting the UK's AI ecosystem and the future of artificial intelligence.
