McKinsey & Company, the strategy behemoth, has been making waves lately with its pronouncements on the future of AI, particularly within the banking sector. But a closer look at their analysis reveals a consulting firm caught between selling a transformative vision and hedging its bets, all while quietly shifting its own business model to capitalize on the AI gold rush.
McKinsey's Global Banking Annual Review 2025, as highlighted by the Financial Times, offers a prime example of this tightrope walk. The report dances around definitive predictions, presenting multiple scenarios for AI adoption in banking. On one hand, AI could "revolutionize everything," replacing customer-facing roles at a staggering 100:1 bot-to-human ratio. On the other, it might "never fully scale," remaining a tool for knowledge workers while your grandparents embrace AI banking agents (a visual I find both terrifying and hilarious).
The consultancy bravely assigns the highest probability to a middle-ground scenario, but even that lacks conviction. The FT writer rightfully points out the questionable math—the numbers don't quite add up to 100%. Is this a simple oversight, or a deliberate attempt to obfuscate the underlying uncertainty? (My experience suggests the latter is often the case.)
This lack of a clear base case is telling. McKinsey wants to be seen as forward-thinking, but it also doesn't want to be wrong. It’s like a weather forecast that predicts sun, rain, and snow, all at the same time. Sure, you've covered all the bases, but you haven't actually told me what to expect when I walk out the door.
While McKinsey hedges its bets on the future impact of AI on its clients, it's making a much bolder bet on AI's impact on its own business. According to Business Insider, McKinsey is "really rethinking the nature of the work that we do," moving away from traditional fee-for-scope models towards performance-based arrangements. About a quarter of McKinsey’s global fees now come from this outcome-based pricing model. Clients are essentially saying, "Here's the outcome we'd like to get to," and McKinsey's fee is contingent on delivering that performance. This shift is driven, in part, by AI transformation projects.
This is a significant change. Consultants have traditionally billed clients based on the hours and resources spent on a project. Now, they're putting their own skin in the game, aligning their incentives with the client's success. It's a clever move, but it also raises some questions. How are these "outcomes" precisely measured? What happens when unforeseen circumstances derail a project? And, perhaps most importantly, does this new model incentivize McKinsey to oversell the potential benefits of AI in order to secure lucrative performance-based contracts?
I've looked at hundreds of these filings, and this particular shift is unusual. It suggests a fundamental change in how McKinsey sees its role—not just as an advisor, but as a partner in driving tangible results. This is a high-risk, high-reward strategy. If McKinsey can successfully leverage AI to deliver measurable improvements for its clients, it stands to reap enormous profits. But if it fails to deliver, it could face reputational damage and financial losses.
The shift towards AI isn't just about pricing; it's also about the type of work McKinsey is doing. "Straight strategy advice" now accounts for less than 20% of the company's work. Instead, clients are turning to McKinsey for "deep implementation expertise" and multi-year transformation projects.
This is where the "double-edged sword" of AI comes into play. McKinsey is selling AI solutions, but it's also selling the expertise to implement those solutions. It's a classic "picks and shovels" strategy—profiting from the AI gold rush regardless of whether the gold actually exists. As AI is reshaping how McKinsey makes money, the consultancy is betting big on this new approach.
Meanwhile, as McKinsey pushes AI solutions, CFOs are taking a more cautious approach to budgeting. According to a CFO Dive article, the general sentiment shaping this year's budget season is a conservative one. CFOs are focused on "protecting the downside" in the face of economic uncertainty, geopolitical risks, and the challenges of budgeting for AI.
This creates a fascinating tension. On one hand, McKinsey is selling CFOs the promise of AI-driven efficiency and growth. On the other hand, CFOs are wary of the hype and are prioritizing cost control and risk management. It's like a car salesman trying to convince you to buy a sports car while you're worried about rising gas prices.
Carmody notes that CFOs are scrutinizing investments more carefully, prioritizing "must-spend" versus "nice-to-have" items. This suggests that McKinsey may face increased resistance to its AI pitches, particularly if it can't demonstrate a clear and measurable return on investment. According to CFOs are reaching for downside budget protections, McKinsey exec says, this is a key trend shaping financial decisions.
McKinsey's AI pronouncements should be viewed with a healthy dose of skepticism. The consultancy is clearly invested in the AI narrative, both for its clients and for its own bottom line. While AI undoubtedly holds potential, the reality is far more complex and uncertain than the hype suggests. CFOs are right to be cautious, and investors should demand more concrete evidence before buying into the AI revolution.
McKinsey's AI Playbook: Revenue vs. Reality
2025-11-21