Auto‑finance fraud surging

Auto-finance fraud is no longer a niche problem buried in subprime lending. It is now large enough that industry trackers expect fraud exposure to top $10 billion in 2026, up about 8.7% from the prior year, after Point Predictive estimated a record $9.2 billion of exposure for 2025 based on 2024 activity (autofinancenews.net, pointpredictive.com). That number sounds abstract until you see what sits inside it: fake borrowers, inflated incomes, washed credit files, and loans that look clean at origination but rot almost immediately once they hit servicing. The striking part is that old-fashioned identity theft is not the center of gravity here. Point Predictive says first-party fraud now makes up 69% of auto-lending fraud risk, meaning the borrower often uses their own name while lying about income, employment, or intent to repay (pointpredictive.com). Income and employment misrepresentation alone account for 43% of total fraud risk, or about $3.9 billion (pointpredictive.com). That matters because these loans do not always look like fraud at first. They look like credit losses. By the time the borrower misses payments, the lender’s fraud team, credit team, and audit team may already be arguing about what happened. That internal confusion is part of the story, not a side note. Auto Finance News reported on April 1 that lenders often review the same file and reach different conclusions about whether a bad document reflects fraud, normal credit risk, or a control failure inside the lender’s own process (autofinancenews.net). The result is delay. Fraud signals get treated as paperwork issues. Credit losses get mispriced. Real errors get mixed with intentional deception. In a market where digital approvals can happen in minutes, that lag is expensive, because the borrower only has to get through origination once. That is why synthetic identity fraud has become so dangerous in auto lending. A synthetic identity is built by combining real and fake personal data into a borrower who does not truly exist, then nurturing that profile until it can qualify for a large loan. Equifax said in March that among super-prime auto accounts, borrowers with high synthetic-risk scores had a 90-plus-day delinquency rate of 7.9%, versus just 0.3% for lower-risk accounts, and the delinquent balances were larger too (equifax.com). TransUnion found something even stranger in October 2025: fraud losses in auto were 21 times higher than in credit cards and six times higher than in unsecured personal loans, with especially large losses showing up in prime and super-prime tiers that are supposed to be the safest (newsroom.transunion.com). Credit washing makes that distortion worse. The scheme is simple enough to spread on social media. Consumers are told to file disputes or false identity-theft claims to temporarily remove negative tradelines, push up their scores, and qualify for financing they otherwise would not get. Point Predictive said signs of credit washing jumped 162% in its 2025 report (pointpredictive.com). Auto Finance News reported on April 2 that the CFPB added extra screening and disclosures to its complaint database after a surge in complaint volume, and the head of the Consumer Data Industry Association said credit-repair companies were using social media to encourage false claims of identity theft to scrub credit reports (autofinancenews.net). The Federal Trade Commission still tells consumers to dispute real errors directly with furnishers and bureaus, which is a very different thing from inventing fraud to erase legitimate debt (consumer.ftc.gov). Now AI is helping fraudsters scale all of this. Point Predictive said conversations in criminal Telegram channels about AI and deepfakes for fraud rose 644% from 2023 to 2024, including tools for fake IDs and synthetic identities (pointpredictive.com). LexisNexis reported in March that synthetic identity fraud was the fastest-growing fraud type globally in 2025, with an eight-fold year-over-year increase, while malicious bot attacks rose 59% and “agentic” traffic surged 450% over the year (risk.lexisnexis.com). Auto lenders do not need every one of those bots to target car loans directly. They only need a few of them to automate logins, generate cleaner fake documents, and push more doctored applications into systems that still treat origination and servicing as separate worlds.