\t\t\t\t\tTable of Contents<\/p>\n
Key Property Data Used for Fraud Analysis<\/p>\n
Valuation Anomalies Are the Biggest Red Flag<\/p>\n
Undisclosed Real Estate <\/p>\n
GeoSpatial Data Show Inconsistencies<\/p>\n
Beyond Mortgage Fraud<\/p>\n
Frequently Asked Questions (FAQ)<\/p>\n
Artificial intelligence (AI) has the speed and power to perform data analysis on layered datasets. Layering data, combined with instant analysis, identifies suspicious information far faster than manpowered systems. Here\u2019s a look at what type of property data are used by financial institutions in their fight against fraud, and how AI pinpoints red flags for cases needing further scrutiny.<\/p>\n
Key Property Data Used for Fraud Analysis<\/p>\n
Fraud rarely shows up as a single obvious red flag. For financial institutions and mortgage lenders, identifying fraud is a process that requires finding patterns and multiple anomalies. These anomalies are often detected only when layers of data are examined and cross-checked.<\/p>\n
What type of data are best for analysis? Typically, property valuation data are the cornerstone of AI analysis conducted by financial institutions<\/a>, but ownership, financial, and geolocation data are all critical components.<\/p>\n Property Values<\/p>\n Mortgage fraud perpetrators often inflate property values. Automated valuation models<\/a> (AVMs) are computer-generated estimates based on mathematical models, tax records, and recent sales data, and they play a major role in finding inflated valuations.<\/p>\n These AI-powered models reveal outliers that could indicate a fraudulent property price. Mortgage lenders typically have teams dedicated to investigating any inconsistencies flagged by the AVM results and when submitted documentation does not align with tax records and sales data.<\/p>\n These teams also monitor appraisers to make sure they are not submitting numbers biased on skewed information. Red flags uncovered from further investigation could include the following:<\/p>\n Discrepancies in square footage or number of rooms in the appraisal, application, and compared to public records. Ownership Data<\/p>\n Lending institutions and mortgage lenders<\/a> also look at ownership activity. Fraudulent property flippers typically seek rapid transactions regarding renovated homes. They inflate home values by colluding with appraisers and fill out loan documents with false information. The rapid resales often leave lenders with outstanding loans, sometimes credited to fake borrowers, that exceed the property\u2019s worth.<\/p>\n Lenders use AI to compare ownership<\/a> and occupation data from reliable, public sources with the information provided on loan applications to spot red flags.<\/p>\n Financial Data<\/p>\n Ownership data are often layered with financial data<\/a>, such as borrower income, credit behavior, and transaction patterns, to find anomalies. If information on a loan application seems suspicious, with additional research, a lender can see where an honest mistake is the culprit, not fraud or intentional misrepresentation.<\/p>\n Valuation Anomalies Are the Biggest Red Flag<\/p>\n Researchers from the University of Texas found that a surge in home prices following the COVID-19 pandemic was partly driven by Paycheck Protection Program (PPP) loan fraud. House prices were inflated in cities like Chicago, New Orleans, and Atlanta, where fraud rates approached 30%. In some cases, home valuations increased by 25% in Texas.<\/p>\n Undisclosed Real Estate <\/p>\n According to the National Mortgage Application Fraud Risk Index \u2014 compiled from millions of U.S. mortgage applications \u2014 the category of undisclosed real estate fraud (which includes undisclosed debt, occupancy misrepresentation, and hidden derogatory credit events) increased by 9.1% year-over-year.<\/p>\n Lenders protect themselves from this fraud by comparing buyer income data to the value of the property. If data show a buyer\u2019s income to be less than stated on the application, that could be evidence of misrepresentation of income. If they do not disclose debt, again that\u2019s a red flag.<\/p>\n Data and AI can provide alerts that a primary or a second home will not be occupied as disclosed. Perhaps an owner-occupied property has a rental listing or a primary residence t has a different tax mailing address.<\/p>\n GeoSpatial Data Show Inconsistencies <\/p>\n Geospatia data<\/a> are valuable to confirm reported square footage on properties. Additionally, financial institutions use geospatial tools to compare the size of the land parcel reported by the borrower to the actual size of the land parcel to find inconsistencies in reporting.<\/p>\n Beyond Mortgage Fraud<\/p>\n It\u2019s not just mortgage lenders and financial institutions that use property data and the power of AI to fight fraud; the Internal Revenue Service (IRS) and the Department of Justice (DOJ) use it to find evidence of tax evasion and to bring criminal charges.<\/p>\n The IRS and the DOJ are increasingly leveraging artificial AI and property data to combat tax evasion, money laundering, and collusion in real estate.<\/p>\n These agencies use machine learning to sift through vast datasets\u2014including property records, ownership databases, and financial transactions.\u00a0 Examples of red flags they look for are mismatched income, undisclosed offshore property, or suspicious rent-setting behavior.<\/p>\n The Internal Revenue\u2019s Use of AI and Property Data<\/p>\n The IRS uses AI to enhance audit selection, focusing on high-wealth individuals, large partnerships, and complex corporate structures.<\/p>\n For example, under the Centralized Partnership Audit Regime (CPAR), the IRS uses AI to target large partnerships (over $10 billion in assets) and those with over $10 million in assets, hedge funds, private equity, and real estate investment firms to find discrepancies.\u00a0 AI programs look for anomalies in record-keeping for complex depreciation, 1031 exchanges, and \u201creal estate professional\u201d status.<\/p>\n The IRS compares reported income against third-party data, including property transactions and land records, to detect disparities that suggest unreported income. The IRS also looks at lifestyle and asset data, such as high-value real estate acquisitions, both domestic and foreign, that do not align with reported income to identify potential tax evasion.<\/p>\n The Department of Justice\u2019s Use of Property Data<\/p>\n The DOJ uses AI to detect fraud patterns often in collaboration with the Treasury Department. For example, the Consolidated Asset Tracking System (CATS)\u00a0uses AI to track real estate and other property seized by federal law enforcement agencies during forfeiture. Some of these properties are found to be involved in crimes or purchased with illicit funds.<\/p>\n Data most used by the DOJ are<\/p>\n non-public rental data, such as rental amounts and pricing strategies AI combined with property data are the engine and the fuel powering fraud detection. Not limited to mortgage lenders and financial institutions, these tools are used by investors, insurers, and government entities to identify false data created and presented by fraudsters.<\/p>\n Contact ATTOM<\/a> data to gain access to industry-leading property data for fraud management.<\/p>\n Frequently Asked Questions (FAQ)<\/p>\n How do financial institutions and mortgage lenders use AI and data to detect mortgage fraud?<\/p>\n
\nAn appraisal that uses comparable sales that are located relatively far from the subject property, are a different type of property type, or not in a similar condition.
\nA valuation that shows a significant unexplained sudden increase in the property\u2019s value when no significant renovations were made.
\nA borrower who is in a hurry and putting pressure on the appraiser to meet a specific valuation.
\nAn appraised value that does not sync with the local market trends.<\/p>\n
\nrecords on property seized during investigations
\ngeolocation data
\nfinancial data for anti-money laundering (AML) investigations. AI can highlight rapid money transfers that might indicate money laundering.
\nProperty ownership data that might hide the true source of illicit funds<\/p>\n