In March 2025, FISD published its Alternative Data Council Buy-Side Perspective: Understanding Investment Firms’ Use of Licensed Data with GenAI. The paper is a thoughtful and constructive contribution to one of the most important licensing and governance discussions now taking shape across capital markets. It is especially useful because it replaces broad assumptions and reactive positioning with a more practical framework for understanding how investment firms are actually using licensed data with GenAI technologies.

From our perspective at PostSig, the paper gets several important things right. It correctly distinguishes between public AI tools, enterprise deployments, and private or proprietary environments, all of which carry materially different implications from a security, confidentiality, and licensing standpoint. It also usefully separates inference, fine-tuning, and full model training rather than treating all AI-related activity as a single category. Just as importantly, it grounds the discussion in the FISD derived data framework, particularly the practical tests of reverse engineering and functional substitution. Those distinctions create a solid foundation for a more precise and commercially workable conversation.

Read more here