Many\u00a0of the tens of millions of Americans shopping for deals today have decided to spend more than they can afford. According to the National Retail Federation, holiday spending is projected to top $1 trillion for the first time, with the average person planning to spend about $890.<\/p>\n
As many as 66\u00a0percent\u00a0of Americans expect to borrow for holiday shopping, and about a quarter\u00a0anticipate\u00a0taking on $900 or more in holiday debt.\u00a0Many holiday borrowers will carry that debt for months on credit cards and buy now\/pay later plans\u2014and, for the most vulnerable, payday loans.\u00a0<\/p>\n
For payday lenders, this holiday season promises to be different: many have deployed AI to speed up approvals and improve targeting.\u00a0And some of their customers have had their precarious finances damaged<\/a> by the longest government shutdown in U.S. history.<\/p>\n Common Knowledge\u00a0<\/p>\n Payday loans have long been marketed as quick fixes for financial emergencies.\u00a0In reality, they\u00a0are among the most expensive forms of credit. The typical payday loan carries an APR around 300\u00a0to\u00a0400\u00a0percent\u2014for example, a $100 loan with a $15 fee for two weeks equates to\u00a0nearly\u00a0391% APR.\u00a0In some permissive states, average rates\u00a0exceed\u00a0600\u00a0percent; Texas has been reported around 664\u00a0percent.\u00a0While\u00a0a number of\u00a0states cap interest at 36\u00a0percent, many still allow triple-digit APRs.\u00a0\u00a0<\/p>\n Roughly 12\u00a0million Americans take out payday loans each year<\/a>, or about 4\u00a0to\u00a05\u00a0percent\u00a0of adults.\u00a0Annual volume exceeds 20 million loans,\u00a0totaling\u00a0$8.6 billion in principal and $2.4 billion in\u00a0fees. Defaults\u00a0are common: about 1 in 5 borrowers ultimately default on a payday loan, according to federal research, and so-called “rent-a-bank” high-cost lenders have reported loss rates averaging around\u00a050\u00a0percent\u00a0in public filings\u2014versus 2\u00a0to\u00a03\u00a0percent\u00a0charge-offs for typical bank consumer loans.<\/p>\n Who are these borrowers?\u00a0<\/p>\n Most earn less than $40,000 a year;\u202f\u00a0Younger\u00a0millennials (25\u201334) account for about 30\u00a0percent\u00a0of borrowers, with Gen Z adults (18\u201324) growing fastest;African Americans and Latinos are disproportionately represented;Many lack\u00a0a college degree;\u202f\u00a0Single renters dominate the market;\u202f\u00a0Southern states and urban areas see the highest usage.\u202f\u00a0<\/p>\n The social consequences are profound. Payday loans often trap borrowers in cycles of reborrowing\u2014over 80\u00a0percent\u00a0of new loans are taken within 14 days of\u00a0a previous\u00a0loan.\u00a0Repeat debit attempts by online payday lenders have been shown to trigger overdraft fees and even account closures for many borrowers.\u00a0Transparency is another issue: borrowers rarely understand why they were approved or denied, and regulators\u00a0have warned lenders that “black-box” models must still generate specific, understandable adverse-action reasons under federal law.\u00a0<\/p>\n And now, artificial intelligence is promising to amplify these problems.\u00a0<\/p>\n Uncommon Knowledge\u00a0<\/p>\n Artificial intelligence is revolutionizing small-dollar lending. What was once a manual, paperwork-heavy process is increasingly automated and decisioned in seconds.\u00a0AI-driven platforms in this space\u2014from loan marketplaces like\u00a0RadCred\u00a0(which uses AI-based matching and soft-pull prequalification) to cash-advance apps such as\u00a0EarnIn, Dave, and Brigit\u2014promise approvals and transfers in minutes by\u00a0analyzing\u00a0income stability, bank-account transaction\u00a0patterns\u00a0and other alternative signals beyond FICO.\u00a0The result? Faster access to cash for borrowers and lower operational costs for lenders.\u202f\u00a0<\/p>\n Speed is the headline feature.\u00a0RadCred\u00a0markets “apply in 60 seconds” and instant AI matching to lenders.\u00a0EarnIn\u2019s\u00a0Lightning Speed transfers arrive within\u00a0around\u00a030 minutes (for a fee).\u00a0Brigit advertises Express Delivery “within 20 minutes.”\u00a0Dave offers Express disbursement to debit for a percentage fee or free ACH by the next day.\u00a0By contrast, many storefront and online payday lenders still fund the next business day via ACH (or\u00a0same-day\u00a0to debit in some cases), meaning access can vary from minutes to a day or more.\u00a0For consumers facing urgent expenses\u2014holiday gifts, travel, or medical bills\u2014instant cash offers instant relief.\u202f\u00a0<\/p>\n But AI\u2019s role\u00a0isn\u2019t\u00a0just speed.\u00a0By\u00a0leveraging\u00a0alternative data, AI can approve borrowers who lack conventional credit histories, theoretically expanding financial inclusion. The companies argue this reduces bias and democratizes access to credit. Consumer advocates counter that proxy variables (like location) can encode bias even when models exclude protected traits.\u00a0<\/p>\n The holiday surge magnifies these risks. Lenders actively market “Christmas loans” and other seasonal credit as applications rise through the fourth quarter.\u00a0<\/p>\n For an unknown number of payday borrowers, that strain\u00a0probably increased\u00a0during the government shutdown that ended in November after a record 43 days,\u00a0with\u00a0federal employees, welfare recipients, and others relying on government funding faced an unexpected cash crunch.\u00a0On\u00a0SoLo\u00a0Funds\u2014a peer-to-peer microlending platform\u2014loan requests jumped 775\u00a0percent\u00a0amid the shutdown\/SNAP pause.\u00a0It\u2019s\u00a0unclear what role, if any,\u00a0AI-powered payday lenders played in shutdown-related borrowing.\u00a0But they are\u00a0structurally\u00a0primed to profit from future crises with approvals at scale, targeted\u00a0marketing\u00a0and surge pricing\u2014think Uber fares in a snowstorm. By Black Friday 2026<\/a>,\u00a0we\u2019ll\u00a0know what the AI arms race has done to America\u2019s most vulnerable holiday shoppers.\u00a0<\/p>\n