Artificial intelligence is giving multifamily owners and operators new tools to forecast delinquencies, prevent them early, engage late tenants more effectively, and reduce reliance on third-party collection agencies.
CRE Daily reported in July of this year that “the share of on-time rent payments in independently operated rental units dropped to 83.6 percent in July 2025 — a 20 basis point (bps) decline from June. That figure marks a 209 bps year-over-year decrease and continues a troubling 24-month streak of declining performance.”
At the same time, GlobeSt. published its 2025 State of Renter Delinquency and Default which among other findings, revealed that multifamily operators have been slow to embrace innovative tools. Seventy-six percent of respondents had not yet explored AI, even though one-third said their existing tools left room for improvement in rent mitigation.
Something has to give.
This Insights by Blueprint research report into the intersection of AI and rent remediation aims to educate operators who are ready for a new way to improve collections and perhaps, turn chronically late payers into engaged community members.
Building the Case for AI Collections
How AI Rent Collection Works
Timing, customization and scale are the primary benefits of implementing an AI solution for collecting past due rent and preventing late payments. These systems work in concert with existing PMS and related digital payment processors and lease administration systems, core sources of the data from which it initiates action.
The systems train on tenant data pulled from the APIs of an existing tech-stack—lease terms, unit type, payment histories, maintenance requests, communications, amenity access. This stage depends on how well and how long the datasets have been managed. Extensive field mapping and cleanup (date formats, naming structures, rent amounts, internal coding, fee definitions, etc.) is often necessary, but with at least a couple of years of data, the next phase runs much smoother.
Once ingested for processing, machine learning kicks in to identify trends and spot anomalies, working with staff to set rules for triggering preventative and reactive tactics tailored to each late-payer scenario.
Those workflows then learn from a series of feedback loops, measuring outcomes and honing data. Better models become highly intuitive, recognizing seasonal impacts on payments, which unit types are usually late, and which apps tenants prefer.
Humans remain in the loop when exceptional instances arise, like legal issues or sudden health concerns. The AI offers recommendations and provides context for further remediation.
The majority of these systems rely on email and text messaging that offer personalized reminders, initiate conversations and drive connections to payment tools. Some will also send automated calls or ringless-voicemail.
Outreach responses feed a backend interface that tracks rent metrics, collection insights, and intervention options, depending on the vendor, as well as a suite of actions to intervene or edit workflows. Two-way integration should be commonplace, and decisions will need to be made about the frequency of data transfers. Real-time syncing works best for operators with frequent delinquencies or where fast follow-up historically delivers better results. Batch processing is fine if performed regularly but know that the data will not be as fresh.
Benefits of AI rent collection
Artificial intelligence captures payments at the earliest opportunity for doing so, minimizing tenant delays and avoidance behaviors. It doesn’t forget to send reminders and generally acts within minutes of rent deadlines, not the next day or two.
AI collection agents also work in mobile and web environments and their intrinsic neutrality can lead to faster and more positive interactions with tenants. At the very least, it alleviates the risk of human frustration and conflict bleeding into the process. It’s not uncommon for burned-out staff to either back away from a difficult collection project or inappropriately ramp up the pressure.
Regardless of communication channel, AI rent collection collates and timestamps documents, text messages, voice interactions, and partial payment amounts for every late payer pursued. In turn, this helps build well-rounded tenant behavior profiles and bolsters potential eviction cases.
For portfolio operators, AI collections offer consistent processes at scale—a powerful argument for full adoption. This makes results easier to forecast and gives organization-wide feedback on how each property performs.
Industry statistics continue to indicate that economic distress is eroding on-time payment trends.
The Consumer Financial Protection Bureau (CFPB) found that “among those who incur a late or non-sufficient funds fee, incurring multiple fees of the same type in a twelve-month period is a frequent occurrence.”
It also found that “around 42 percent of renters incur another late fee in the month immediately following their first late fee, and around 30 percent are still experiencing late fees five months later.”
Human nature being what it is, no technology built can eliminate the risk of tenants paying late. What it absolutely can do, though, is cut the time, stress and costs of apartment operators’ efforts to remedy it.
Forecasting Delinquency with AI
Apartment operators can lean on deep use cases of AI to reduce risk closer to the top of the funnel. Debt collection analytics leverages sentiment analysis, for example, emanating from ongoing communications between tenants and leasing offices in both specific collection scenarios and throughout general tenant experience campaigns.
An AI model can identify financial stress through the integration of external resources for consumer financial data, considering utility bill payments, real-time credit scores and machine learning models used in credit decisions. They also consider ebbs and flows in spending frequency, account balances, job status indicators and even social media content.
This type of analysis is beyond the ability of leasing and accounting staff using traditional methods. An AI can ramp payment reminder cadence as soon as the model sniffs risk, present links to partner rent assistant programs or offer payment plans. It can also apply findings over time and at scale to build applicant personas that can be mapped against new applicants.
Some data used in delinquency forecasting is subject to privacy restrictions and consumer permissions. Most vendors in the space will be able to tell you how deep they go in this regard and it’s best to work with vendors who demonstrate transparency on these issues.
Lastly, and at risk of digressing slightly to further a point, consider that in 2023, “the financial services sector allocated roughly $35 billion towards AI projects. Recent projections estimate that the global AI in finance market is expected to reach $190.33 billion by 2030, growing at a compound annual growth rate (CAGR) of 30.6% from 2024 to 2030.” This was reported by Coherent Solutions, a digital engineering firm strongly tied to the financial sector. And the thrust of that spending?
“The core of artificial intelligence within the financial sector is represented by financial modeling and prediction,” the company said.
Enhancing The Payment Experience
Popular consumer payment providers are common to AI collection efforts. They provide fast and secure methods to get funds and are designed for mobile use cases. Expect to see common fintechs like PayPal, Zelle, Venmo, Stripe, Google Pay and Apple Pay.
Also, major mobile carriers allow payments through native or third-party texting apps, speeding delivery and creating a familiar experience. These platforms are faster than ACH transfers and certainly checks being dropped through a slot.
In general, the more mobile-friendly the collection effort, the better the returns. The good news on this front is that every solution in this space is going to meet that requirement.
While payments via ACH continue to make the most sense on a broad scale, both financially and from the standpoint of integration with PMSs, operators need to have modern consumer-facing payment processes at the ready when time is of the essence.
Tailored and Respectful Communications
Vendors in the space can deliver campaigns in multiple languages, adding to the bespoke manner in which AI is able to operate. They can also adjust timing based on when a recipient most often responds, as well as adjust the format. Email and text for one, voicemails for another.
Respecting personal boundaries builds trust and increases confidence that payment requests reach residents.
Over time, AI can learn which outreach methods are more likely to lead to successful resolutions and when needed, use those opportunities to offer direct connections to state and federal public assistance programs for struggling tenants.
Faster Response and Recovery
AI’s ability to deliver calm, frequent follow-ups is one of the best tools for ensuring late rent gets collected. The ability to trigger late notices directly to tenants within minutes, if needed, can remedy a host of escalation risks, from the accumulation of fees to the increased odds of a tenant completely skipping town. Responses from the tenant can also reveal their intent sooner, helping determine if vacancy is going to follow.
One such industry provider, Colleen AI (now an Entrata product) highlights its AI’s ability to react quickly to late payers.
“Speed is especially key in recovery as the chances of collecting drop 16 percent every 30 days an account remains delinquent with larger balances dropping even more from 40 percent to a mere 5 percent recovery rate,” Colleen CEO Itamar Roth in a December 2023 press release, in which the company stated it reduces the rate of unpaid rent by 40 percent. A year later, that rate was at 50 percent.
Fraud Detection and Prevention
Financial services application Plaid reported that “The share of businesses reporting ACH credit fraud has increased by 6 percent since 2021 and more than half of organizations with revenue less than $1B were unable to recover funds lost from payments fraud attacks.”
AI collection agents can stir together sentiment analysis, risk profiles and financial anomalies to detect fraudulent payment attempts. Examples include inconsistent payment promises, ACH fraud, questionable documentation, new accounts, and obscure payment platforms—issues that usually take longer for property managers to spot.
Balancing AI with the Human Touch
Buzz is a company that supplies multifamily owners with “hyper-realistic conversational AI” to engage with late-paying tenants, typically by phone.
“Because it sounds like a real person, opt-out rates are super low,” Buzz Co-Founder and CEO Jake Lisby said at the 2025 Multifamily Innovation Summit. “If they know it’s a pre-written message your opt-out rates go through the roof. Phone calls are the most effective way to collect money, it shakes them into paying attention, and we send a text message along with that. Nine times out of ten they respond with a text, but the phone call gets their attention.”
Three apartment communities that use Lisby’s AI improved their collection revenue by 35.25, 40 and 63.70 percent respectively. In total, the three properties saved 288 hours of labor per month, the company said.
Tenant-reward programs are also useful tools to personally engage tenants and alleviate lost rent.
Research from AI rent services company Domuso found that 40 percent of multifamily communities use some form of rewards program with 93 percent of that group finding them beneficial.
Smart property owners are finding ways to pass off reward management to their AI, making tracking easier, speeding content delivery, and drawing more attention to on-time payments.
Rentgrata, an AI-based applicant and resident experience company (recently acquired by industry colleague Opiniion) begins its sentiment analysis at the top of the resident marketing funnel, citing
Other considerations
Data Access and Sharing Challenges
Data collection is paramount for AI’s impact on rent collections to be fully realized.
While regional and local employment trends, cost-of-living indexes and other market-specific microeconomic data on individuals are more widespread. Challenges arise when AI must integrate with existing systems like lease documents, payment records, or maintenance requests.
Larger software providers can be hesitant to provide APIs or other data gateways that help power an AI’s needs for delinquency forecasting, proactive engagement and skip-tracing triggers. Property owners should clearly determine to what extent their existing tech-stack is willing to share data with an AI provider before committing to an AI solution for collections.
Consumer-Permissioned Data in AI
Consumer permissioning of financial data allows consumers to fence off specific sections of their digital financial history, making available only what’s needed to approve an application.
This new level of granularity does create challenges but it also opens doors to new, alternative credit histories, such as previous leasing history, consumer payment app profiles, even social media usage. Information gathered in this way is more worth having, as its precision leads to smarter, faster decisions and more comprehensive tenant profiles to be used in application acceptances, delinquency forecasting and when needed, arrears management.
Financial services application Plaid is the most common name in this space, and offers services directly to property owners.
Any party requesting information has additional verification hurdles to clear. Informed consent must be explicitly granted and the requesting party’s use case transparent and specific.
There is the risk of bias in AI-based credit decisioning and delinquency forecasting. The complexity of the machine learning engines and algorithms are often difficult for the average consumer to understand. Do they, for example, apply more weight to a short-term lease or to the absence of a savings account?
This can translate into trouble for the property owner with an AI that “assumes” one tenant over another is more likely to miss a payment. Ask about this to ensure AI software providers are clear on their AI’s ability to act without prejudice.
Nevertheless, when it comes to building digital applicant and tenant financial pictures, these new fintech-inspired practices are well worth understanding by the tech-forward property owner.
Compliance and TCPA Considerations
There is also risk of TCPA (Telephone Consumer Protection Act) rules altering how businesses can reach customers.
Strict permission-based rules were intended to become law in January of 2025 but a lack of public awareness resulted in a year’s delay. Property owners should remain aware of how it may impact AI-generated outreach to current residents and those who remain past-due. It’s safe to say that tenancy and the legality of a lease meets the standard of “pre-existing relationship,” and thereby would justify AI-calling and communication. Clearly this is an issue best left to legal departments.
–
There’s little doubt that increasing economic pressure will both increase the value of collections automation as well as the need for property management teams to operate with fewer resources. AI’s rapid move into arrears management couldn’t be better timed.
Still, everything related to AI requires strict attention to the pace of its evolution. Property owners and management teams should remain diligent in their efforts to understand its role in all aspects of the operation. There will be consolidation and rapid changes in product quality, cost and integration resources.
Despite its rapid learning, AI shouldn’t be trusted to operate alone—yet. Don’t assume this is an auto-pilot scenario. There will be a role for manual intervention and oversight for some time, especially when collection efforts become more complex or begin to raise legal questions.
In July, a CRE Daily rent collections report said that “the broader forecast full-payment rate — which factors in late payments and typical late-pay behavior — fell to 93.4%. That’s the weakest reading since December 2020, at the height of the pandemic-era rental disruption.”
It’s clear that now is the time for multifamily operators to modernize how they prevent and collect late rent.
-Craig C. Rowe
