When I talk to CRE firms about AI, I hear the same line over and over: “We’re going to wait and see how this plays out before we really invest.” On paper, that sounds prudent. No one wants to chase every shiny tool. In practice, “waiting to see” on AI for CRE is not neutral. You are quietly paying for it every day in ways that do not show up on a P&L line, but absolutely show up in your pipeline, your people, and your positioning. This is what that bill actually looks like. 1. You Lose The Compounding Effect Of Small Experiments AI is not one big switch you flip. It is dozens of small, messy experiments that stack over time. When you push those experiments off: You still underwrite every deal You still scrub every rent roll You still write every IC memo by hand You just do it without building a repeatable AI for CRE playbook along the way. Firms that started a year ago did not magically automate everything. What they did was: Try AI on the rent roll parsing, figure out what worked Try AI on OM summarization, refine the prompts Try AI on IC memo drafts, tune the voice Every quarter, the hit rate improves. The quality of prompts improves. The list of things they trust AI with grows. If you are “waiting to see,” you are not standing still next to them. You are falling a year behind on a compounding learning curve that is very hard to catch once client expectations move. 2. Your Best People Quietly Build Their Own Stack Without You Analysts, associates, and even some principals are already using AI on their own machines. If the firm is not providing: Clear guidance on what tools to use Standard workflows for common tasks Guardrails around data and confidentiality then your smartest people start building their own AI stack in private. They: Paste rent rolls into random chatbots Use general models to summarize legal language Store prompts and snippets in personal notebooks You pay the risk. They get the productivity. The organization learns almost nothing. The cost of “waiting” here is simple: You do not see what is working You cannot standardize or scale the good stuff You cannot protect the firm from the bad stuff By the time you decide to get serious, the talent that has been quietly doing AI work for you is either frustrated that leadership is behind or has already been poached by a firm that wants that skill set. 3. You Keep Funding Your Competitors’ Data Advantage Every deal you touch creates data exhaust . Underwriting models DD memos Negotiation trails Rent rolls and operating statements If you are not feeding that into an AI for CRE engine that you control, that data just sits in shared drives and inboxes. Meanwhile, any competitor who is even moderately serious about AI is: Indexing every deal they touch Letting AI learn from wins, losses, and passes Turning that history into: Better screening rules Smarter sensitivities Faster “no” and sharper “yes” The hidden cost is that you keep doing the hard, expensive part of the work, and they reap the pattern recognition benefits. Your work becomes raw material for someone else’s learning curve. 4. You Negotiate Against AI With No AI Of Your Own We already talked about this in another piece, but it is worth repeating here. In almost every meaningful negotiation now, the other side has: AI summarizing your emails AI comparing your mark‑ups to prior deals AI drafting counters and talking points If your position is: “We’ll adopt AI after the market settles down,” what you are really saying is: “We are comfortable negotiating against people who outgun us on preparation, memory, and pattern recognition.” The cost is not visible on day one. It shows up as: Slightly worse covenants Slightly tighter timelines Slightly riskier language that you “did not notice” in a long document Each one looks small. Across a portfolio, it is material. 5. You Signal To Clients And Investors That You Are A Follower Clients and LPs do not need you to be a tech company. They do expect you to be: Curious Data‑driven Willing to test new tools where the risk is controlled When they ask about AI and you say: “We are waiting to see what the rest of the market does,” what they actually hear is: “We are more comfortable copying than leading.” “We will probably be late to the next obvious improvement too.” That might be fine if you are selling pure yield, but if you are selling: Strategy Process Edge then “wait and see” is not a neutral story. It directly contradicts your positioning. 6. You Keep Paying For Manual Work That Could Already Be 50–80% Automated Some parts of CRE work are not ready for full automation. Fine. Other parts are frankly not good to still be doing by hand: Renaming and sorting OM pages Re‑keying rent rolls into Excel Manually pulling basic tax and ownership data Drafting the first pass of an IC memo from scratch When firms say “we are not ready for AI yet,” what they often mean is: “We do not want to change how work is done.” The cost is straightforward: You pay fully loade