Every firm that has introduced AI to its team has encountered some version of the same moment. Someone in the room, often someone experienced, often someone whose judgment the firm genuinely relies on, pushes back. Maybe it is direct. Maybe it is quieter than that, a skepticism that shows up in the way they engage with the tools or the way they talk about them to the people around them. Either way, leadership feels the friction and has to decide what to do with it.
How that moment gets handled matters more than most firms realize. Not just for the person pushing back, but for everyone watching to see what happens next.
The instinct in a lot of organizations is to treat resistance as a problem to be overcome. Get the skeptics on board, neutralize the friction, and move forward. That instinct is understandable and usually wrong. The people who push back on AI in contracting firms are rarely pushing back on technology in the abstract. They are pushing back on something specific, and that something is usually worth understanding before it gets dismissed.
Sometimes the concern is about quality control. A senior estimator who has spent twenty years building reliable numbers has a legitimate interest in whether AI outputs can be trusted in a context where being wrong costs real money. That is not resistance. That is professional judgment, and it deserves a real answer rather than a reassurance.
Sometimes the concern is about relevance. Someone who has built their value inside the firm around a specific kind of expertise, the person who knows how to read a complex spec, who understands the quirks of a particular GC’s contracts, who carries institutional knowledge that nobody else has, reasonably wonders what their role looks like if AI starts doing what they do. That concern is also legitimate, and telling someone their expertise is safe without showing them specifically how it fits into an AI-assisted workflow is not going to move them.
Sometimes the concern is simpler than either of those. People are busy. Learning a new tool takes time and cognitive energy that feels like it is coming out of the same account as everything else they are responsible for. If the firm has not made space for that learning, the resistance is a reasonable response to an unreasonable expectation.
The firms that navigate this well tend to do a few things differently from the ones that struggle.
They separate the people who are skeptical from the people who are genuinely unwilling, because those are different situations that require different responses. Skepticism is an invitation to make a better case. Someone who is genuinely unwilling to engage regardless of what they see or hear is a management issue that AI did not create.
They involve the skeptics early rather than presenting AI as a finished decision. The estimator who is worried about output quality is often exactly the right person to help define what good output looks like and where human review needs to stay in the loop. Giving that person a role in shaping how AI gets used in their area of the business turns potential friction into genuine ownership. People who help build something are far more likely to use it and defend it than people who have it handed to them.
They are honest about what AI does not do well, because credibility with a skeptical team depends on not overselling. If leadership presents AI as a solution to every problem and the team finds the edges quickly, which they will, the credibility of the whole initiative takes a hit. A more honest framing, here is where this tool produces real value, here is where your judgment still has to lead, tends to land better with experienced people who have seen enough vendor promises to be appropriately cautious.
And they recognize that the pace of adoption will not be uniform, and that is acceptable. Some people will move fast. Some will take longer. Forcing everyone onto the same timeline in the name of consistency tends to produce compliance without genuine use, which is worse than a slower and more uneven adoption that actually sticks.
The goal is not a firm where everyone is equally enthusiastic about AI. That is not a realistic picture of how people work. The goal is a firm where AI is available, understood well enough to be useful, and supported by leadership in a way that makes engagement feel worthwhile rather than obligatory.
The people who push back hardest are sometimes the ones who, once they find a use case that connects to their specific work, become the most effective and credible advocates for the tools inside the team. That outcome is worth working toward. It starts with taking the pushback seriously rather than trying to manage it away.