Forbes Warns AI Giants Shifting to Revenue-Sharing Models

Forbes is right to raise the alarm: AI pricing is entering a new phase, and it could reshape how marketers, SaaS operators, and growth teams think about margins.

For the last few years, AI was sold in familiar ways. Companies paid flat monthly subscriptions, bought team seats, or used APIs priced by tokens and requests. It wasn’t perfect, but it was predictable. You could forecast it, budget for it, and treat AI like a tool you rented.

That model is starting to change.

From software fee to revenue participation

What Forbes highlighted is more than a pricing tweak. It points to a structural shift in how AI vendors want to get paid. Instead of charging only for access, some companies are moving toward pricing based on outcomes, results, or a share of the value their AI helps create.

We’re already seeing early examples. Intercom’s Fin AI charges per resolved conversation. Salesforce has positioned Agentforce around conversations handled. Elsewhere, AI vendors are charging based on completed workflows, recovered revenue, successful claims, or performance-based percentages.

Then came OpenAI CFO Sarah Friar’s comments outlining a broader monetization strategy that includes licensing, IP-based agreements, and outcome-based pricing. That’s when this stopped looking like a niche experiment and started looking like a roadmap.

If the biggest AI companies are signaling that they want to share in the value created, it’s reasonable to assume this model will spread.

Why AI giants are making this move

The vendor logic is easy to understand.

AI is expensive to run. The economics are different from traditional SaaS. Old-school software could enjoy very high margins once the product was built. AI doesn’t work that way. Compute costs stay variable, infrastructure demands keep climbing, and the race to scale models is massively capital-intensive.

From an AI provider’s perspective, subscriptions and token fees may not be enough. If their models help close deals, resolve support tickets, optimize campaigns, or contribute to product discovery, they want a larger share of that upside.

On paper, outcome-based pricing sounds reasonable: pay for results, not just activity. Lower upfront risk. Better alignment.

In practice, it’s much messier.

The real risk for marketers and operators

Once AI vendors shift from software providers to revenue participants, your cost structure changes with them.

That matters a lot in marketing.

Imagine an AI platform helping manage paid media and then taking a percentage of ad spend, attributed conversions, or incremental revenue. At first glance, that sounds performance-friendly. Over time, though, that vendor isn’t just powering your workflow. It’s compressing your margin.

Forbes’ warning matters because this won’t stop with one category. It can move into customer support, media buying, lead qualification, e-commerce, creative generation, and attribution tooling. The more embedded AI becomes in revenue operations, the easier it is for vendors to justify taking a cut.

That’s where a simple software bill starts to look more like a toll booth.

Why outcome pricing creates friction

There’s another issue that doesn’t get enough attention: attribution.

Who actually created the outcome?

Was it the model? The human strategist? The sales team? The creative? The offer? The brand? Existing demand? The budget? Usually, it’s all of the above. That makes revenue-sharing and outcome-based deals highly vulnerable to disputes.

Critics have already pointed out the principal-agent problem. The vendor is motivated to define success in ways that maximize its share, while the customer is left trying to audit a black box. Even if the contract looks reasonable on day one, renewals can get much harder once the real bills start showing up.

And that’s before IP questions enter the picture. If AI providers begin tying pricing to discoveries, outputs, or commercially valuable work, businesses will need to think much more carefully about ownership, licensing terms, and downstream usage rights.

What I believe smart teams should do now

The answer isn’t panic. It’s preparation.

• Audit every AI contract. Look for vague terms around outcomes, derivative value, usage overages, attribution, and licensing. If a vendor can later reinterpret what counts as success, that’s a risk.
• Diversify your stack. The more dependent you are on one AI provider, the weaker your negotiating position becomes. Compare commercial models across closed and open providers, not just model quality.
• Insist on cost visibility. If AI is entering your growth engine, you need clear reporting on what you’re paying for and why. Hidden variability is where margin leakage begins.
• Keep humans in the loop where value attribution matters most. The more strategic the work, the more dangerous it is to let pricing formulas decide who deserves credit.
• Separate enablement from ownership. Paying for a powerful system is one thing. Letting a vendor claim an ongoing share of the value your team creates with that system is another.

The bigger shift behind the headline

What Forbes is really pointing to is the end of the AI-as-simple-software era.

As agents become more autonomous, vendors will try to position them not as tools but as contributors—digital workers, partners, even co-creators. Once that framing takes hold, revenue-sharing becomes much easier to pitch. If the AI did the work, why shouldn’t the vendor get a percentage?

That argument will appeal to boards and investors because it dramatically expands revenue. For customers, especially marketers and operators managing tight ROAS targets, it introduces a new layer of cost unpredictability exactly where financial discipline matters most.

That’s why this trend deserves serious attention now, before it becomes standard contract language.

FAQ

What is outcome-based AI pricing?

Outcome-based AI pricing means vendors charge based on results rather than simple access or usage. That could include fees tied to resolved support cases, completed workflows, recovered revenue, or a percentage of performance.

Why are AI vendors moving to revenue-sharing models?

AI is expensive to build and run. Variable compute costs, infrastructure demands, and pressure from investors are pushing vendors to look for pricing models that capture more of the value their systems create.

What’s the biggest risk for marketers?

The biggest risk is margin compression. If AI vendors take a percentage of outcomes tied to media buying, lead generation, or conversions, software costs can start eating into profitability in ways that are harder to predict and control.

Why is attribution such a problem?

Because business outcomes rarely come from one source. Results usually reflect a mix of strategy, creative, brand strength, budget, demand, and execution. That makes it difficult to assign clean credit to an AI vendor without creating disputes.

How should companies respond?

Review contracts closely, diversify vendors, demand better cost transparency, and be cautious about any agreement that gives a provider an ongoing claim on the value your team creates.

Conclusion

AI should improve profitability, not quietly siphon it away. As revenue-sharing models spread, businesses need tighter control over attribution, spend efficiency, and actual performance economics. If you want a clearer way to manage return and protect margin while AI pricing models get more aggressive, build around systems that keep ROAS visible and actionable, and ROAS Suite is a smart place to start.