There’s a study from MIT worth paying attention to if you’re a building owner who pays professional services fees. It found that employees at firms that had adopted AI were using it daily, but weren’t reporting it. They were getting work done faster and keeping the difference.

Think about what that means for you.

The billable hour was always a strange deal

The professional services model in construction and design has a fundamental misalignment at its core. The consultant gets paid for time spent. You get charged for the hours it takes to produce a deliverable. If they can produce that deliverable in less time, there’s no mechanism to pass the savings to you, in fact, there’s an incentive to not reduce the hours at all.

This has always been true. What changes now is the scale of the gap. When AI can draft an engineering analysis in minutes that previously took hours, or review a specification document faster than any junior staff member, the question of who captures that efficiency gain becomes very pointed.

If the answer is “the firm,” you’re paying 2025 rates for what is increasingly 2025-speed work. The economics of professional services are being renegotiated whether the parties involved acknowledge it or not.

The shadow AI economy

What the MIT study describes is already playing out across every firm you hire. Staff are using AI tools to move faster: on emails, on analysis, on documentation. Most aren’t telling their employers how much faster. And employers aren’t passing the savings to clients.

This isn’t malicious. It’s just how incentive structures work. Firms don’t voluntarily cut their billing rate. But it does mean you’re in an era where the relationship between hours worked and value delivered has fundamentally changed, even if the billing structure hasn’t.

The firms that will earn your loyalty over the next five years are the ones honest enough to restructure how they charge for the work. Value-based agreements, outcome-based fees, shared savings arrangements. These models align your interests with your consultants’ in a way that time-and-materials billing no longer does.

What this means for your contracts

You have more leverage here than most owners use. A few practical shifts worth considering:

Ask about AI usage directly. When you interview firms, ask how they’re using AI and where it reduces their cost to deliver. The honest firms will have a real answer. The evasive ones will tell you it’s complicated.

Structure fees around deliverables, not hours. A project schedule review shouldn’t cost $8,000 because it took two people eight hours. It should cost what it’s worth to you to have the review done accurately and on time. Negotiate from there.

Require transparency on where AI-generated content appears in your deliverables. Not because AI-assisted work is bad, it can be excellent, but because you should know what you’re getting and what you’re paying for.

The deeper shift

The billable hour problem is a symptom of a larger restructuring happening in professional services across every industry. As AI handles more of the analytical and documentation work that used to require human time, the value of expertise concentrates in judgment, relationships, and accountability, not in the production of deliverables.

An engineer who can tell you which structural system best fits your project constraints, explain the tradeoffs clearly, and take accountability for the outcome is worth paying for. The junior hours spent producing the calculation set to document that conclusion are increasingly something you shouldn’t have to pay for at 2019 rates.

The construction industry will restructure around this. The question for owners is whether you shape that restructuring or just absorb the costs of the transition while the firms figure it out on your projects.

You’re already paying for AI speed. Make sure you’re getting the benefit.