The afternoon briefing.

AI Copyright. The Atlantic has launched a searchable database of music used to train AI models, revealing millions of tracks in datasets downloaded thousands of times. This initiative aims to bring transparency to the often-opaque process of AI training, allowing artists and the public to investigate potential copyright infringement. The move highlights growing concerns over intellectual property rights as AI models increasingly rely on vast amounts of existing creative works. This comes amid reports of direct plagiarism, where AI was used to relaunch an author's stolen book.

AI Talent. A significant shift in AI talent sees Nobel laureate John Jumper departing Google DeepMind for rival Anthropic, underscoring the intense competition for top researchers in the frontier AI space. Meanwhile, Microsoft's CSO acknowledges the human struggle to keep pace with AI advancements, warning of a narrowing window to understand and control these rapidly evolving systems. OpenAI's Codex is expanding its capabilities with a "Record & Replay" feature, allowing it to learn and automate complex workflows after a single demonstration.

AI Market & Regulation. Concerns are mounting over the financial stability of the AI sector, with an NYU finance professor warning that a potential AI crash could be more severe than the dot-com bust due to heavy debt-financed infrastructure. This comes as tech investors increasingly monitor the bond market, given the substantial capital expenditure required for AI buildouts. Regulatory bodies are also grappling with defining AI-generated content, as seen in the EU's struggle to classify deepfakes in retail advertising, posing challenges for implementing transparency rules.