In credit markets, the edge often goes first to the firm with the least legacy to defend.
Electronic execution gained share fastest among firms that could redesign around it from the start. Direct lending grew into a trillion-dollar market built by managers who did not need to migrate an existing syndicated business. In both cases, the advantage went first to whoever moved with the fewest internal switching costs.
The same question is now forming around AI.
A fund launched today can design its research infrastructure around AI from the first day. Not a tool added to an existing process, but the foundation of how coverage gets done. What I have seen in practice is small teams screening more issuers, extracting covenant terms consistently across documents, and moving from raw materials to an investment view with less manual friction. The productivity gain is not uniform, but it is real and it compounds.
Established managers are not excluded from this. The barrier is not awareness or capability. It is internal: change management, budget allocation, and a conversation about what the existing research model was built to do. Some are approaching it at the strategy or fund level rather than firm-wide, which may be the faster path.
For LPs diligencing credit funds today, AI operating model is increasingly part of what belongs alongside team and track record.