In February 2026, OpenAI announced its Frontier Alliance: multiyear partnerships with Accenture, BCG, Capgemini, and McKinsey to help enterprises deploy its Frontier agent platform. Each firm is standing up dedicated practice groups, getting teams certified on OpenAI technology, and gaining early access to product roadmaps and technical resources in return.
Capco — where I’m starting as a Principal Consultant in a few weeks — has its own OpenAI services partnership, announced in November 2025, focused on financial services and energy. Not part of the Frontier Alliance by name, but structurally similar: priority access to OpenAI’s technology, solution architects, and training resources, in exchange for building and selling OpenAI-powered solutions to clients.
I have a conflict of interest in writing this. I’m about to work at one of these firms. So take what follows with appropriate salt.
Here’s the question that’s been sitting in the back of my head: if your consulting firm has a commercial partnership with OpenAI — certified practitioners, shared roadmap access, co-selling arrangements — what does that do to the quality of advice they give clients about AI platforms?
The honest answer is: it creates a structural tension that most firms won’t surface in their engagement letters.
Consulting firms have long had vendor relationships. The Big Four audit-adjacent arms built entire practices on SAP and Oracle implementations. IBM Global Services and IBM product were famously intertwined for decades. The pattern isn’t new. What’s different now is the speed and the stakes. The AI platform decisions companies are making in 2026 will probably be sticky for five to seven years — not because switching is technically hard (it mostly isn’t, at the API layer) but because workflow redesign, change management, and institutional muscle memory are expensive to redo. Getting recommended into the wrong platform isn’t just a licensing cost; it shapes what your workforce learns to do.
So when a consulting firm has material incentives to deploy OpenAI — certified engineers who need billable Frontier projects to justify their investment, practice leaders who get roadmap access contingent on deal flow, commercial relationships whose health depends on client adoption — the question isn’t whether individual consultants are corrupt. They’re not. The question is whether the institutional incentive structure reliably surfaces “actually, a different platform fits your needs better” as an option.
I don’t think it reliably does. Not because the advice is dishonest, but because the frame narrows before the analysis even starts. The engagement is scoped around deployment, not selection. By the time a client is talking to Accenture or Capco about AI transformation, the vendor question is often pre-answered by the relationship.
The counterargument — and it’s a real one — is that these firms still have reputational incentives to deliver results. A bad implementation is worse for the firm’s brand than losing one platform deal. And for firms like McKinsey and BCG, whose leverage comes from being trusted on hard calls, compromising advice quality is existentially costly.
That’s true. I believe it, actually. But “we won’t give you actively bad advice” is a lower bar than “we’ll tell you when a different vendor would serve you better.” The Alliance structure, as publicly described, doesn’t obviously create incentives for the latter.
OpenAI declined to disclose the financial terms of the partnerships. That opacity matters. Whether there’s revenue sharing, volume-based certification tiers, or softer incentives like preferential roadmap access tied to deal counts — the mechanics shape the advice whether or not anyone in the room acknowledges it.
What should a client do with this? A few things, none of them revolutionary:
Ask your advisor directly: what platforms did you evaluate, and what’s your commercial relationship with each? The answer will tell you something.
Separate strategy from implementation. If the same firm is recommending a platform and implementing it, that’s two bets on the same relationship. Consider whether an independent strategy review — from a firm with no implementation revenue at stake — would change the framing.
Read the partnership press releases. They’re written for investor relations, not clients, but they’re honest about what the firms are optimising for. “Building teams certified on OpenAI technology” means those teams need OpenAI projects to justify the investment.
None of this means don’t use these firms. They employ genuinely smart people with real domain expertise. But structural conflicts of interest don’t disappear because the people inside them have good intentions. Knowing the shape of the incentives is just part of getting good advice.
I’m going to be inside one of these structures shortly. I expect I’ll be thinking about this a lot.