Debt by Data

In January 2025, the loan agency BMG Money announced a major expansion of its mission to reach “overlooked” borrowers. Workers excluded from traditional financial institutions via the “archaic method” of credit scoring could now access affordable loans, according to a press release. To find those new borrowers, BMG Money would rely on a $100 million startup called Argyle. 

Argyle hoovers incredibly granular, real-time worker data from payroll platforms at companies like Starbucks, Walmart, and Amazon, as well as Uber, Lyft, and DoorDash. It then sells that data to financial institutions like BMG Money, DailyPay, or other companies like insurers and vehicle lenders, who target those same workers. Though unknown to most people — even those whose companies use it — Argyle claims to be the leader in an expanding niche of the data economy (Truework, Konfir, and Mastercard’s Finicity have also emerged in recent years). These companies say they’re facilitating a “democratization of credit,” but what they’re really doing is classic predatory lending with a high-tech sheen. Delving into how Argyle works is an opportunity to think about the intersection of data privacy and labor rights—and how we can turn vague concerns about privacy into material demands.

Argyle’s info-harvest works like this: A Walmart employee in need of extra cash might turn to a company like DailyPay, which provides “earned wage access” — or instant access to part of their wages ahead of payday. Argyle creates a live connection between the worker’s payroll account and DailyPay, allowing lenders to see if the worker skips a shift, takes unpaid leave, or quits. Argyle also collects information across over 170 categories, including the worker’s shift start and end times, the frequency and duration of their breaks, and their promotion history. The company can see how many rides an Uber driver completes as well as their location and duration, canceled rides, tips and bonuses, acceptance rate, timeliness, and customer ratings. 

“These granular details feed into algorithmic risk models,” says Wilneida Negrón, Director of Research and Policy at the labor advocacy organization Coworker, “effectively shaping a personalized profile of each worker’s ‘risk.’” By delivering creditors real-time flows of comprehensive worker data, Argyle helps them identify those borrowers who are just risky enough to be profitable and helps to usher new borrowers into the subprime fold.

Argyle’s ongoing ability to scrape massive amounts of data from vulnerable groups is due to its deft deflections of criticism. In 2021, Vice found that a company seemingly backed by Argyle sent workers emails offering to pay them $500 for their payroll login credentials, in what looked like a classic phishing scheme. A 2022 WIRED article described a worker who had unknowingly granted Argyle access to his payroll account when applying for a payday loan — he’d missed the tiny print that let him know the process was “powered by Argyle” and didn’t realize what all the terms and conditions box he’d clicked allowed them to do. 

Argyle quickly jumped into action following the coverage, updating their website to include a section stating “user consent and data privacy underpin everything we do. We’re working to help consumers exercise real ownership over their digital information.” Today, the company is eager to tout its commitment to personal privacy.

But the idea that Argyle is actually providing workers “real ownership” of their data is dubious, if not laughable. I spoke with researcher Samantha Dalal, whose current study on algorithmically-determined wages relies on worker data collected via Argyle. When she messaged the 240-person WhatsApp group for a Colorado gig workers’ union, many of whom had authorized Argyle’s access to their payrolls for her study, not one person had even heard of the company.

Argyle isn’t the only company to promise to protect your data even as it profits off it: Apple’s new ad campaign, with the slogan “Privacy. That’s iPhone,” is case in point. But perhaps even more than Notes app or Instagram, data brokers like Argyle highlight the political necessity of defining demands for privacy in specific, material terms, rather than accepting the going definition of “consent.” Multiple studies have shown that most people skip over terms and conditions pop-ups. The average worker has no legal counsel, nor the time or the expertise to review such policies. They also have no real ability to bargain over the proposed conditions. And, most importantly, they might need whatever is on the other side of the data transaction (like a loan) too much to contest or refuse. What does agreeing to Argyle’s terms and conditions mean under those workers’ conditions? 

By focusing on individuals’ control over their data, we may have missed the forest for the trees. While the erosion of our informational autonomy is worrisome, the material consequences of data extraction are far more concerning. A Reddit search for DailyPay returns hundreds of distressed posts from borrowers with titles like “How do you get out of the DailyPay loop?” or “DailyPay is the devil…” As Negrón explains, these “financial tools are marketed as worker-friendly, but often operate as exploitative stopgaps that trap people in cycles of liquidity stress.” Reports published by the Center for Responsible Lending back up this statement. In the case of earned wage access providers like DailyPay, which make up some of Argyle’s customers, the reports found that one in three of their borrowers are trapped in perilous debt cycles and pay up to 300 percent in interest rates. 

A myopic emphasis on individual privacy also obscures how companies like Argyle impact collective action. As organizers I spoke with informed me, Argyle now holds the datasets that worker groups need, too. As a result, they may be forced to turn to Argyle to purchase back their own data. “In a very short time, Argyle has added another layer of proprietary technology that further locks workers out of their data,” says Negrón. This dynamic may create constraints on the efficacy of organizing efforts. By becoming beholden to Argyle for data access, organizers help to reinforce the data monopoly of gig economy companies.

These contradictions are why scholar Veena Dubal discourages efforts to “reclaim” workers’ data through intermediaries like Argyle. Instead, she calls for organizing and legislative action against the harmful practices this data enables — like algorithmically-determined wages. In the case of Argyle, our focus should not be on improving its terms and conditions contracts but on addressing how its data collection practices impact workers’ material reality. What we really need is an outright ban on the sale of worker data and stringent protections against workplace surveillance. Rather than an individual right, privacy must be a matter of class solidarity. 

Eliza McCullough is a researcher and writer interested in technology, surveillance, and labor rights. Her work explores the algorithmic management of gig workers and labor rights for data enrichment workers. She leads research on labor issues at Partnership on AI.