Subject: Q1 2027 AI licensing plan: fund proof before scale
Preview text: A practical planning agenda for content owners, with three scenarios and explicit conditions for further investment.
AI Training Data Licensing Research | Q1 2027 planning outlook, prepared 4 November 2026
The planning position
Approve a sequence of decisions rather than a single archive-monetisation forecast. Set a ceiling on each stage of investigation, identify the evidence needed to release the next budget, and retain the ability to stop.
This outlook does not predict licensing market growth or assign probabilities to outcomes. Its purpose is to help a publisher or specialist information owner enter the quarter with a testable plan.
The infrastructure context
IAB Tech Lab says CoMP version 1 was finalised on 28 April 2026. Its stated scope includes technical communication and packaging, while explicitly excluding a licensing marketplace and the definition of economic models. Read the standards overview.
Planning implication: separate a technical implementation decision from a commercial appointment. A successful integration can answer whether your systems exchange the required information; it cannot, by itself, answer whether the proposed rights or economics are acceptable.
For structured information owners, the 30 September Gateway announcement also illustrates a possible request-based distribution route. Read the announcement. Its existence is a reason to test fit, not a reason to redesign a profitable existing product immediately.
Three scenarios to prepare for
| Scenario — planning assumption | Evidence that would support it | Recommended response | Condition for stopping or changing course |
|---|---|---|---|
| A qualified buyer needs a narrow, differentiated cohort | Named decision owner; written use; evaluation and contracting path | Release an approved representative sample and cost the defined delivery | Use expands beyond cleared rights or acceptance remains undefined |
| Interest is real but commercial requirements remain incomplete | Technical discussion and follow-up, without confirmed terms | Continue inexpensive qualification; cap preparation spend | Repeated requests for broader access without a clearer decision |
| Activity is largely promotional | Generic marketplace pitch; no archive-specific requirement or usable references | Retain options and revisit only after new evidence | Upfront investment or exclusivity is demanded without adequate support |
These are decision templates, not descriptions of named suppliers. A buyer can move between scenarios as evidence changes.
The quarter's operating agenda
First decision — what can we credibly offer? Nominate a lead cohort, not the entire archive. Produce a brief covering subject, dates, modality, update needs, known permissions and unresolved restrictions. Have legal and data colleagues identify the smallest evaluation package that can answer the commercial question.
Second decision — is the route worth testing? Compare direct discussions, representation and infrastructure on the work each actually performs. Ask who originates demand, who contracts, who prepares data, who receives money and who owes the owner a report. Remove duplicated charges or unassigned responsibilities before judging the headline fee.
Third decision — what did the test prove? Use agreed acceptance criteria. Record whether the sample improved the buyer's intended workflow, what preparation remained necessary, and whether the owner could verify the resulting obligations. Separate positive technical feedback from a commercial commitment.
Fourth decision — should we expand? Bring finance an assumption-led model that includes preparation, counsel, updates, deductions, timing and the alternatives constrained by exclusivity. Review it against the evidence obtained, not the enthusiasm generated by the pilot.
What the board should receive
Provide a short recommendation with four outcomes available: proceed, proceed with conditions, hold, or decline. Attach the evidence register, unresolved rights, expected spend at risk, accountable owners and the next review date. Show which numbers are signed terms, supplier estimates or internal scenarios.
Ask the board to approve only the next justified commitment. An owner that declines an unsupported mandate has preserved an option; that is a legitimate result of diligence even when no licence is signed.
Next step: choose one live archive decision and use the demand-before-preparation decision gates in the do-it-yourself plan before committing the quarter's engineering capacity.