"We protect your privacy" — but the trip log says hi Open the data export tool inside the rideshare app of your choice. Most people are surprised to find: Every pickup and drop-off, down to GPS coordinates and second-precision. Every "looked at the app and didn't book" event, including the addresses you typed. The driver's identifier and vehicle plate, by trip. Phone-derived signals (battery, Wi-Fi state, motion sensors). These records routinely span 5 to 7 years for active accounts, even in jurisdictions where the published policy claims shorter periods. What the public record shows Dutch DPA fine against Uber (Aug 2024) — €290M for transferring European driver data to the US without an adequate transfer mechanism after the Privacy Shield was invalidated. The DPA press release in the citations is the primary source for the scope, the legal basis, and the data categories at issue. EFF on rideshare data sharing with police (2018) — documents the gap between the published policies and the real flow of trip data into law-enforcement requests, and is still the clearest framing of why mobility logs are a civil- liberties issue. NYC TLC public trip-record dataset — the city publishes anonymised trip records that researchers (most notably Anthony Tockar in 2014) have repeatedly shown to be partially re-identifiable when joined with other datasets. The TLC page in the citations is the canonical source. Uber's current Privacy Notice — the linked notice is the primary source for what Uber currently retains, why, and for how long. How to shrink your trip file Use the in-app data export to see what is held. Both Uber and Lyft publish a self-service deletion flow under "privacy." Delete trip history older than 12 months. It is one button each side. Most riders never use it. Disable "always" location at the OS level. "Only while using" is enough; "always" lets the app build a 24/7 mobility profile. Don't link payment cards you also use for sensitive merchants. The cross-merchant graph is one of the most resold signals. The bigger pattern Mobility is one of the easiest ways to deanonymise a person. Four timestamped GPS points over 6 months are enough to uniquely identify 95% of people in a metropolitan area (the De Montjoye / Nature 2013 result, replicated multiple times since). Rideshare apps are not the only mobility log — but they are one of the few you can actually shrink today, in two minutes, with a delete button. Use it.