Should data portability be infrastructure for agent-driven markets?

The CMA argues that data portability is not a nice-to-have feature for agentic commerce but a structural requirement for competitive markets to function once agents mediate consumer transactions. The CMA states portability "is not simply a convenience: it is a structural condition for dynamic markets," because "agentic systems are only as effective as the information they can access, yet today consumer data remains fragmented across businesses, sectors and formats."

The reasoning is that an agent's usefulness depends on what it can see: an agent that cannot pull a consumer's preferences, history, or context from wherever that data currently sits cannot act well on the consumer's behalf, and a consumer who cannot move that data between agents cannot switch providers freely. The CMA frames "smart data initiatives" as the enabler that would let agents "act on up to date, user authorised information" and "operate across providers without repeated manual intervention." Absent that, the CMA warns of a specific competitive failure mode: "lock-in" to incumbent agentic ecosystems, where switching costs become a barrier that "undermines competition and the potential for agentic innovation to drive household prosperity and growth."

This is a competition-policy complement to Will brands succeed by building their own AI shopping agents?, which treats the agent as a decision layer brands must win over; the CMA's portability argument explains why that decision layer could ossify into a small number of incumbent ecosystems if consumer data cannot move between them. It also runs parallel to Why do consumers delay adopting agentic search?: both describe mechanisms by which a platform's structural position — control over data, or control over the search layer — can outpace what is good for the consumer unless something, portability or catching-up incentives, corrects it.

The CMA's excerpt makes a structural argument, not an empirical one: it does not cite data on current switching costs, market concentration among agent providers, or evidence that lock-in is already occurring, since consumer-facing agentic deployment is still described as limited. The claim functions as a forward-looking policy recommendation — that data mobility infrastructure should be built now, as a precondition for competitive agentic markets — rather than a finding that lock-in has already materialized.