Back to blog
7 min read

Nine New AI Business Models, and the Three I Would Bet On

  • Business Models
  • Industry Trends
  • AI Agents

Every major consulting firm has now published its own framework for how AI changes the shape of a business, not just its cost base. The frameworks differ in naming, but they converge on the same underlying idea: the companies that win are not the ones using AI to do the same work cheaper. They are the ones using it to sell something they could not sell before.

I spend my working life running P&L for enterprise accounts in gaming, hospitality, fintech, and consumer credit, so I read these frameworks the way I read a proposal: which parts would actually close, and which parts are slideware.

The three I would actually bet on

  • Outcome based delivery, where a client pays for a result an agent helps produce, not for a headcount running the process
  • AI augmented managed services, where the managed service itself gets faster and cheaper to run because agents handle the repeatable layer, and that margin gets reinvested into the relationship
  • Embedded expertise products, where domain knowledge that used to live in a person's head gets packaged into an agent based tool a client can use between engagements

The ones I would be careful with

Fully autonomous, agent run business lines sound exciting in a keynote and fall apart the first time a regulator, an auditor, or an angry customer asks who is accountable. In gaming and fintech specifically, that question gets asked early and often. Any business model that cannot name a human accountable for an agent's decision is not ready for those industries yet, no matter how clean the framework slide looks.

My rule of thumb when I evaluate one of these frameworks for a client: if the model cannot survive being explained to a regulator in one sentence, it is a pitch deck, not a strategy.