On September 20, 2026, WIRED's Aarian Marshall reported on an academic workshop convened that same day in Gothenburg, Sweden asking a question the mobility industry has been carefully not asking: how do you design a driverless vehicle so people don't have sex in it? The session — "Love In Traffic: Physical and Digital Design Against Consensual Sexual Activity Inside Public and Shared Automated Vehicles" — ran as Workshop W1 at AutomotiveUI 2026, the conference's official programme listing organisers Alexandros Rouchitsas, Ignacio Alvarez, Soyeon Kim, Chen Peng and Pavlo Bazilinskyy.

Why It Matters

This is what it looks like when sexual behaviour becomes a formal engineering requirement in a mainstream technology sector. AutomotiveUI is an ACM-affiliated conference with automotive OEMs in the room; a workshop on designing against sexual activity means detection and deterrence specs are being drafted for vehicles that will carry millions of passengers. Whatever comes out of it will be built into hardware and software long before there is public debate about it. The surveillance implications deserve more scrutiny than the headline invites. A vehicle that can detect sexual activity is, necessarily, a vehicle running continuous interior behavioural monitoring — cameras, occupancy and motion sensing, possibly audio — with an on-board classifier making judgments about what passengers are doing with their bodies. That is a substantially more invasive sensing stack than ride-hailing has today, justified by a use case that is easy to agree with and generalises to everything else. Rouchitsas's blinking-light proposal is a real-time behavioural judgment rendered by a car. The privacy questions the connected-device segment has spent a decade arguing about — what intimate data gets collected, who sees it, how long it is retained — arrive here fully formed and almost entirely undiscussed. There is also a legitimate accessibility and equity tension buried in the organisers' own reasoning. They want automated mobility to be "equitable, serious, societally minded" and explicitly cite benefits for the elderly, children and people with disabilities. Fair. But a robotaxi's private, unsupervised cabin is also genuinely useful to people whose living situations afford no privacy at all, and interior monitoring calibrated to detect intimacy will not distinguish between consensual adults and anything else — it will simply flag bodies. Whatever design principles emerge from Gothenburg, the industry should be asked to publish the detection method alongside them.

The research method was, by the organisers' own account, unglamorous. Bazilinskyy, an assistant professor at Eindhoven University of Technology who has spent over a decade researching AI-driven human-robot interaction, told WIRED he found himself typing "public transport" into the amateur section of a major adult tube site. It returned 297 hits. "People are hard to predict," he said. The workshop abstract makes the same point more formally: prior research has covered either fear of sexual harassment in public transport or consensual activity in private cars framed as distracted driving, while "evidence from adult content platforms suggests that consensual sexual activity inside public transport is trending."

The stated concern is neither moral nor prurient. The abstract argues that in shared automated vehicles, "enclosure, shared use, and the absence of an authority figure can warp perceptions of accountability and intervention potential," with consequences for other passengers' comfort and for public acceptance of the technology overall. Rouchitsas, an engineering psychologist, was blunt with WIRED about the framing: he and his colleagues are "taking the side of the parents, the elderly—we are not the cool people here," and his fear is that self-driving gets co-opted by "the partiers, the drug users." Or, as he put it: "You know 'Bang Bus'? We don't want that."

The proposed interventions are early and speculative — Rouchitsas floated transparent vehicles, or a light that blinks when the vehicle's sensors detect behaviour it disapproves of. The four-hour workshop ran at Natrium, Gothenburg University, at a conference sponsored in part by BMW and Google, in a city that is home to Volvo Cars and sensing supplier SmartEye. Organisers intend to publish a paper based on the results. Bazilinskyy's broader argument is that sex is the attention-grabbing case for a more general design problem: "Prevent sex and maybe you'll also prevent robbery, bullying, and homicide."

Sources

Update — 2026-09-21

Initial entry — story first created.