UK Policy Reversal on AI Training Data
In a pivotal development, the UK has backed away from its earlier stance that AI training on copyrighted materials falls under fair dealing exceptions. This reversal means AI developers now require explicit licenses or permissions to use protected music in model training. According to Music Business Worldwide (Source 1), the decision responds to intense lobbying from labels and publishers concerned about devaluing catalogs. The change brings UK law in line with stricter EU regimes under the AI Act and ongoing US debates. Industry groups like the BPI hailed it as essential for sustainable AI innovation that respects creators. This could deter unlicensed scraping by firms like OpenAI and Stability AI, potentially spurring licensing deals. Rights holders anticipate clearer revenue streams from AI outputs mimicking their works.
BMG's Lawsuit Against Anthropic Escalates AI Copyright Wars
Music major BMG has sued AI developer Anthropic, alleging infringement through unauthorized use of copyrighted lyrics and compositions in training its Claude models. Billboard reports (Source 2) the suit seeks damages and injunctions, building on similar actions by Universal and Sony. BMG claims Anthropic's scraping of lyrics sites violated fair use defenses. This case tests whether AI training constitutes transformative use or direct copying. With Anthropic's rapid growth, outcomes could set precedents for tech-music negotiations. The lawsuit underscores publishers' push for opt-in licensing amid AI's data hunger. Analysts predict settlements leading to revenue-sharing models, mirroring deals with Spotify and YouTube.
$8M AI Music Streaming Fraud Guilty Plea
A defendant has pleaded guilty in a massive $8 million scheme using AI-generated tracks to inflate streams on platforms like Spotify. Per Billboard (Source 3), the fraud involved bots simulating millions of plays to claim royalties from PROs and DSPs. Authorities uncovered fake artists and metadata manipulation exploiting payout thresholds. This marks one of the largest convictions in streaming fraud history, highlighting vulnerabilities to AI tools. The DOJ emphasized deterrence against synthetic content abusing legitimate systems. Platforms are ramping up AI detection, but perps adapt quickly. The case prompts calls for blockchain provenance in music to verify human creation.
IFPI Execs Weigh In on AI Licensing Needs
The IFPI's latest report features executives urging standardized licensing for AI music generation and training. Music Week (Source 4) quotes leaders from Warner, Sony, and indies on balancing innovation with rights protection. They advocate opt-out registries and equitable revenue splits from AI products. Amid UK and US shifts, execs see licensing as inevitable, citing Suno and Udio suits. The report warns unregulated AI risks market flooding with low-quality outputs, eroding artist incomes. Global frameworks like WIPO discussions gain traction. IFPI pushes for transparency in training datasets to enable fair compensation.
Implications for Global Music-AI Regulation
These developments signal tightening global scrutiny on AI's use of music IP. UK's reversal, BMG's suit, fraud convictions, and IFPI advocacy converge on licensing mandates. According to sources, this pressures AI giants to negotiate rather than litigate. Europe leads with DMA rules, while US fair use battles loom in courts. Labels eye collective deals via PROs for efficiency. Creators benefit from emerging royalties on AI covers and stems. However, open-source AI poses enforcement challenges. Policymakers must address deepfakes and voice cloning next, ensuring tech serves art.