OpenAI starts charging some customers only when its AI actually works
Key Points
- OpenAI and other software vendors are testing outcome-based pricing. Large customers pay only once the AI completes a task successfully.
- High operating costs are forcing the industry to rethink. Vendors like Salesforce now tie their prices directly to the revenue gains or cost cuts customers see.
- The hard part of the model is attributing the success. Without clear rules, it's tough to prove that good results really came from the AI.
Software vendors are moving away from fixed subscription fees and billing by usage or by results instead. According to a report, OpenAI is now doing the same, according to The Information.
Over the past few months, OpenAI has started giving some large customers the option to pay only once its AI actually finishes a task, such as handling customer support requests. That's according to The Information, citing a person with direct knowledge of the arrangements. The move hadn't been public before. An OpenAI spokesperson declined to comment.
OpenAI is adopting a model that startups pushed hardest. It's outcome-based pricing, where customers pay only when the AI works. Sierra and Fin, the latter now being acquired by Salesforce for $3.6 billion, charge only for tasks the AI completes without human involvement. The coding assistant Cognition promises corporate customers credits of up to $10 million if the software fails to deliver results worth at least as much as the price they paid. Older vendors like Adobe, HubSpot, and Zendesk are heading the same way.
Why the subscription fee is under pressure
There is some pressure on the AI companies to adopt such measures, as their products are expensive to run, and so far they haven't sped up revenue growth at software companies, according to The Information. At the same time, customers' IT budgets are strained by competing tools like Anthropic's Claude, which recently pushed harder toward billing by usage.
Salesforce now lets companies choose how they pay for its own AI, Agentforce, including individually negotiated contracts tied to how much the AI lifts revenue or cuts costs. Customers want to buy and price in different ways, CEO Marc Benioff said at an investor conference. The point isn't just to say the software closed a certain number of calls and charge two dollars for them. It's to say the software raised revenue by a certain amount, so the customer should pay two dollars because the AI brought in 20 or 40.
Who gets credit becomes the sticking point
The more common the model gets, the trickier it becomes to say whether a cost saving or a closed sale came from the software or from the customer. The payment provider Stripe has already published guidelines on this. A result could come from product changes, marketing campaigns, or seasonality, Stripe notes. Without clear attribution rules, customers could dispute that the vendor drove the success.
Adobe has also said it will bill part of its newly bundled AI suite, CX Enterprise, by the value it creates, such as the number of ad campaigns completed. Adobe didn't name specific prices. Subscription and usage models stay in place for other tools, and existing AI features for photo and video editing are excluded.
AI News Without the Hype – Curated by Humans
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.