This essay was adapted from Data & Society researcher Alex Rosenblat’s keynote speech at re:publica in Berlin, May 7, 2019, which was originally published at Points.

On May 10, Uber will debut

This is the same company that, along with Lyft, insists that it’s not in the transportation business because it’s a technology company. Both companies argue that they are therefore not obliged to provide accessible services under the Americans with Disabilities Act, unlike their competitors in the transportation business. Uber’s ambition to be the future of transportation excludes its potential obligations under existing law. Uber’s logic is part of a wider pattern of using technology to sow doubt in how we name what is easily observable. Facebook, for example, monitors users’ posts to detect indications of suicidal behavior and then stages interventions, such as calling the police, as Natasha Singer reported. However, by refusing to call this “healthcare” or, potentially, “practicing medicine,” Facebook is able to play by a different set of rules, like running experimental health research algorithms on unwitting users.

What I’ve seen through four years of research with Uber drivers across the U.S. and Canada is how much technology disrupts a shared set of facts and understanding. Uber looks like a taxi company, but it sidesteps regulations designed for transportation companies by self-identifying as a technology company. While Uber promotes drivers as entrepreneurs and classifies them as independent contractors, I found that drivers do have a boss–an algorithmic one. The control that emanates from a faceless boss is often hard to pin down, but the effects are evident. For Uber’s drivers, technology is being used to exclude workers from the benefits they would be entitled to as employees, like a minimum wage.

Uber leverages technology to change how we understand their work in the first place. Algorithmic management isn’t necessarily that different from management–Uber still exercises significant control over how drivers behave on the job. That issue of control is at the center of disputes over whether drivers should be classified as independent contractors, or as employees. When drivers from Massachusetts and California sued Uber in a class-action lawsuit alleging they were misclassified as independent contractors, Uber’s lawyers argued in open court that drivers are actually customers of Uber’s technology, just like passengers.

That sounds like just wordplay, but it’s also changing the paradigm. Uber drivers may have to seek redress for workplace concerns, like wage theft, not under labor law but under consumer protection laws that prohibit unfair and deceptive practices. Uber drivers only account for 0.56% of the labor force, but Uber’s cultural impact is far greater because it is introducing paradigm shifts.

Uber constantly advances the culture of technology to suggest that it isn’t what it looks like. And that cultural work is largely successful in the United States, with direct implications for its business model. For example, Uber is either misclassifying millions of drivers, or it is coordinating prices for millions of small, independent firms (“drivers”), which may violate antitrust laws, as economist Marshall Steinbaum observes. Uber can exist with multiple, conflicting truths, but that doesn’t cut both ways. For example, the city of Seattle tried to extend collective bargaining rights to drivers, but the Federal Trade Commission and the U.S. Chamber of Commerce sided with Uber to successfully mitigate the city’s efforts: Independent contractors can’t bargain collectively over pay because it’s inconsistent with antitrust laws and labor laws.

Uber’s technology ideology comes from Silicon Valley, and how that becomes entrenched in law and practice is a microcosm of a larger political battle for power and governance. Technology has changed how people access information, as technology and social media scholar danah boyd argues. Tech platforms, from YouTube (owned by Google) to Facebook, are structured to personalize our experiences as search engines and recommendation systems. These structures are raising new questions about how we arrive at a common set of facts in a networked media environment, especially because the vulnerabilities in that information architecture can be manipulated to shape public knowledge, as researcher Rebecca Lewis demonstrates. At a time when there is more political uncertainty about governance in the U.S., platforms are increasingly recognized as proxies for political power. We appeal to Facebook, Twitter, or Google to regulate the spread of information and ideologies.