A leasing coordinator at a firm managing a few hundred doors opens an application for a two-bedroom unit. It comes with two recent pay stubs, a bank statement and an employer letter with a phone number. The year-to-date totals add up, the withholdings look plausible, and when she calls the number, someone confirms the job and the salary. The lease is signed and the keys are handed over. Three months later the rent stops, the employer line is disconnected, and the firm is explaining to an owner client why their unit is now headed into an eviction.
Larger operators have started to measure how often this happens. In an NMHC and NAA survey of 75 apartment owners and managers, 93% said they had experienced fraud in the previous year. On average, respondents attributed nearly a quarter of their eviction filings to fraudulent applications followed by nonpayment. Falsified income documentation was the most common type of fraud reported, ahead of identity theft and unauthorized occupants.
Source: National Multifamily Housing Council
View data
| Share of respondents | |
|---|---|
| Fake pay stubs or income documents | 84.3% |
| Misrepresented application info | 80.0% |
| Identity theft or fake IDs | 70.0% |
| Unauthorized occupants or subletting | 67.1% |
| Fraudulent checks or payments | 62.9% |
Smaller landlords are seeing the same pattern. The San Francisco Standard reported this month on a Checkr survey in which 70% of San Francisco landlords said they had discovered or suspected application fraud that ended in an eviction, and 28% said a fraudulent tenant had cost them at least $5,000. Bisnow cites an NMHC-affiliated technology group putting the average cost of mitigating a single case at about $15,000. For a firm paid a percentage of collected rent, a single bad approval can cost more than the unit earns it in fees for a long time, and it is the manager the owner holds responsible.
Screening checks the applicant, not the paperwork
Most tenant screening was designed around a credit report, a criminal and eviction search, and an income check. The first two come from third-party databases. The income check usually does not: it is whatever the applicant uploads, typically a handful of PDFs and an employer contact. That asymmetry is the opening. A credit report can confirm that a person exists and has paid bills before, but it says nothing about whether the employer on the pay stub is real or whether the deposits on the bank statement were typed in.
Producing convincing documents has become cheap. Generative tools can lay out a pay stub with consistent withholdings and a matching bank statement in minutes, and the more organized operations go further. Bisnow reports that some fraudsters now register real LLCs and issue pay stubs from them, so the employer exists on paper and someone answers its phone.
The regulatory side pulls in a different direction. Screening is increasingly regulated, and much of that regulation sensibly aims to protect renters from repeated fees and opaque denials. Colorado, for example, requires landlords to accept a portable screening report prepared for the applicant within the previous 30 days and bars charging an application fee when one is provided. Fair housing rules also push managers toward applying identical steps to every applicant, which makes ad hoc extra scrutiny of individual files risky. And once a lease is signed, undoing a bad approval runs through the eviction process. The Standard notes that San Francisco landlords worry about running afoul of the city's rent laws when trying to remove a tenant who got in by fraud.
Where the usual defenses break down
Owners on landlord forums describe a familiar set of countermeasures. The most common is to call the employer, but only after finding the number independently rather than using the one on the application. That works for large employers with an HR line. It is slow and inconclusive for small businesses, contractors and gig workers, and the calls and callbacks can stretch over days while a competing landlord approves the same applicant.
Requesting bank statements so that deposits can be matched to pay stubs is the next layer. In practice it means a person comparing amounts, dates and pay frequencies across several documents by eye, and bank statements can be edited as easily as pay stubs. Screening bundles sold through listing sites and property management software are useful for credit and court records, but most do not examine the income documents at all.
Document forensics vendors exist, built mainly for large multifamily operators. They catch edited PDFs well, but they charge per application, and renters with legitimate scanned or unusual payroll documents complain about being flagged. When that happens, a staff member has to decide anyway. Identity checks have their own gap: Bisnow describes a test by MRI Software in which it bought 200 AI-generated fake IDs, some for as little as $5, and found that optical card readers flagged only 26%.
None of these tools is useless. Each catches a slice of the problem. The difficulty for a small firm is that stitching them together falls to one leasing coordinator who is also showing units, chasing maintenance and answering owner emails, and who has an owner asking why the unit is still vacant.
What has shifted in verification
Two changes make the problem more tractable than it was a few years ago. The first is that more verification can now go to the source. Applicants can increasingly connect a bank or payroll account directly, so income is read from the institution rather than from a PDF they control. Experienced landlords in forum threads increasingly recommend this kind of source-level check over reviewing documents.
The second is that software can now read the documents that remain. Language models can pull figures out of pay stubs, bank statements and offer letters reliably enough to check them against each other: whether year-to-date totals fit the pay frequency, whether net pay matches the deposits, whether the employer's address and phone appear in public records. That reconciliation used to require either a dedicated reviewer or an enterprise vendor contract.
It is worth keeping the counterargument in view. Bisnow quotes a National Consumer Law Center attorney asking whether fraud concerns are being used to justify charging applicants more. More screening is not automatically better, and checks that wrongly reject people with real income carry their own costs, legal and otherwise.
Fitting verification into a small firm's leasing workflow
For a small management firm, the practical gap is less about any single check than about running the same complete set of checks on every file without adding a day to each approval. A custom-built automation can sit between the application inbox and the property management system, extract every figure from the submitted documents, reconcile them, look up employers independently, and give the leasing coordinator a short list of specific inconsistencies with the evidence attached. The decision stays with a person. What changes is that it is made with the cross-checking already done, recorded the same way for every applicant.