Artificial intelligence is rapidly reshaping commercial real estate, and no segment feels the shift more acutely than multifamily. According to new industry data, the share of CRE companies running AI pilots—or planning to—has surged from just 5% in 2023 to 92% today, with many testing multiple initiatives at once.  

As operators race to understand which applications will meaningfully transform leasing, maintenance, asset management, and resident experience, few leaders are as well-positioned to assess the next chapter as Melanie French, CEO of RREAF Residential. With more than 16 years of executive leadership across major operators and responsibility for a portfolio exceeding 30,000 units, French offers a clear-eyed view of how AI will redefine multifamily operations, investment strategy, and talent models in 2026. 

CM: Industry research shows AI adoption in CRE jumping from 5% to 92% in two years. What do you see driving this explosive shift among multifamily operators?  

MF: Expenses continue to rise, in many places at a faster pace than rental rates are rising. We continue to see rising delinquencies leading to an increase in loans being placed in Special Servicing and even into foreclosure.  The business of home rentals must adapt and find ways to remain profitable in these times. With utilities, taxes, insurance and payroll among the highest expenses, it is natural for owners and managers to seek enhancements and programs that will automate recurring administrative task. AI’s progression and growth provide one avenue to do this.  

While adoption rates may be increasing, it remains to be seen whether this foray into automation will be successful, cost effective and accepted by both customers and users. So far, AI launches have not gone as well as anticipated by the various multifamily leaders I regularly speak with. It is improving, but not at as fast a rate as hoped. 

CM: Are you seeing multifamily companies approaching AI as experimentation—or as an operational necessity? 

MF: A bit of both, depending on which AI product you are talking about. The necessity is that we must find ways to advance into a more efficient business. AI can help with this, but for every two success stories where it works well, you will also have at least one to three examples of an AI launch that failed.    

CM: Where do you believe AI will deliver the biggest operational efficiencies by 2026—onsite staffing, maintenance, budgeting, or something else? 

MF: I likely have a different opinion on this than most, as I believe the greatest efficiencies will be around using AI for task and processes that haven’t been really done by site teams. What do I mean by this? It’s great to think that AI will replace work being done by onsite or corporate teams currently, but the real truth is that staffing levels on site have changed very little in the last 30 years, despite the workloads increasing dramatically with technology, computers, reporting and resident expectations increasing.  

Examples of where I see AI making most impact efficiency wise includes an earlier and higher collection of debts owed by former residents, follow up to customers on service request, condensing the 75-page lease document into bite-sized highlights that residents can understand, conducting audits of leases, assessment of service request for improved efficiency and identifying preventive measures and completing more frequent resident touchpoints.  

All of these are areas and processes that we like to say our on-site teams do as policies require, but the truth is that these are not being completed well currently. Teams just don’t have time, or don’t make time to do these tasks, even though they should be. AI will help do all of this, and that in itself will help to improve asset operational efficiencies and ultimately revenue that comes from improved resident and employee satisfaction.  

CM: Which property management functions are most ripe for automation today, and which still require high-touch human decision-making? 

MF: My prior response covers the automation factor. High-human touch will still be required to build a sense of community which drives resident “nesting” and satisfaction. Most people still find it annoying to reach automated systems when they call a business, yet it is becoming the norm in our industry. If you have a choice to reach a live person on the phone versus a bot, you will likely choose the live person for anything more than the simplest needs.  

Human nature means people still want to feel valued, appreciated, and recognized. AI is becoming better and better at emulating human conversation, but AI still can’t bake cookies and bring them to the office to say thank you to the service person for handling their maintenance ticket. There is a reason the book on identifying a person’s love language was such a huge success. Those human interactions matter. Always have and always will. 

CM: Many CRE companies are running multiple AI pilots at once. What challenges or friction points are you seeing across early implementations? 

MF: We have done this at my firm. Sometimes successfully and sometimes not.  Teams are suffering from technology and change fatigue with the rapidly changing pace of technology. Our biggest headaches are seen with the platforms not being launched well, trained and implemented well, and of course, the fact that they rarely do everything they say they will, nor do integrations with your other already in place systems work well.  

CM: Which areas tend to overpromise—and which consistently deliver value? 

MF: In my experience, everyone we have worked with has touted promises that did not initially pan out. It has felt more like we are helping them build their platform than stepping into one already operating well. We have seen consistent value creation with those focused on online payment portals, collection activities, and on some lead response administrative processes.  

CM: If you had to name one AI application that will define multifamily operations in 2026, what would it be? 

MF: The maintenance functions are ripe for additional automation and improvement. I am hopeful that in 2026, we will see everything from improved predictability in mechanical failures before they happen, the ability to use technology to alleviate things like frozen pipes and sprinkler system failures, while enabling us to find ways to improve the make-ready process and improve the time we spend on the most time consuming service tickets.  

CM: What advice would you give operators who feel behind the curve but want to adopt AI strategically rather than reactively?  

MF: Don’t rush into a program or platform just because someone else is doing it, or just because you like the person leading that platform. Do your advance work in evaluating where your biggest needs are, what problems you are trying to solve for, and then find the tool to solve those needs. Don’t settle. My second piece of hard-learned advice would be to spend more time ensuring your planned launch and implementation is thoughtful, deliberate and over communicated to the users.