WUFT: algorithms cut worker pay

Scheduling software sounds boring. But for a lot of workers, it decides rent money, child care, groceries, and whether a job still feels like a job or just a slot machine. That’s the point of this story. A new NPR report carried by WUFT on May 3 follows LanguageLine Solutions interpreters whose hours got chopped up after the company brought in new scheduling software during a downturn — and whose pay fell with it. ### Who are these workers? They’re phone interpreters — people who jump into hospital visits, court proceedings, city services, and other high-stakes conversations when someone needs language help right now. One of them, Brooklyn-based Haitian Creole-English interpreter Yves Valerus, had what looked like a normal full-time job for about a year and a half: regular weekly hours, a set hourly rate, and benefits. Then 2025 changed the deal. ### What changed at LanguageLine? LanguageLine said demand fell and it started using new scheduling software. After that, workers’ hours became both lower and more fragmented. Valerus told NPR that by the end of 2025 her pay was almost 20% below the prior year. Labor organizers say some coworkers saw paycheck cuts as high as 40%, and some were furloughed. That’s the key thing here — the software did not just rearrange shifts. It changed income. ### Why does fragmentation hurt so much? Because a schedule can look “efficient” on a dashboard while wrecking a worker’s day. A solid eight-hour block is one thing. A few disconnected chunks spread across the day is another. The worker stays tethered to the job, but the paid time shrinks. Valerus described making harder tradeoffs at home, including prioritizing internet over utilities because she works remotely and traveling farther for cheaper groceries. ### Is this just one company’s mess? Not really. It’s part of a bigger shift toward algorithmic management — software that assigns work, measures productivity, sets pay, or all three. Harvard’s Shift Project says unstable scheduling is already a defining feature of service-sector work for millions of people, and its survey data covers more than 200,000 hourly workers in retail, logistics, and more. ### Why do companies use these systems? Basically, they promise tighter labor costs and higher utilization. If software can match staffing minute by minute to predicted demand, managers can avoid “overstaffing.” But the catch is that the cost gets pushed onto workers as uncertainty. Human Rights Watch has argued that algorithmic systems increasingly hire, monitor, compensate, and discipline workers in ways that are opaque and hard to contest. ### Why are unions suddenly central here? Because workers want a say before software rewrites the job. LanguageLine interpreters are organizing with the Communications Workers of America, and New York City officials recently backed their union push. Their complaints are not abstract AI fears. They’re about hours, breaks, wages, and whether the company can quietly change the terms of work through a scheduling system. ### Are lawmakers paying attention? Yes — slowly. Bloomberg Law reported in January that states including California, Colorado, Georgia, and Illinois had proposed guardrails for AI-based compensation decisions, with more efforts returning in 2026. A lot of the debate is still early, but the direction is clear: regulators are starting to treat algorithmic pay and scheduling as labor issues, not just software choices. ### Bottom line? This story is really about who absorbs the volatility. The company gets a more responsive staffing model. The worker gets a shakier paycheck. And once that logic is baked into software, it can spread fast — unless workers, unions, or regulators force the system to answer to something other than efficiency.