Uber tightens driver background-check rules after reports violent offenders cleared

Uber is again discovering the joy of selling “frictionless” mobility while outsourcing the messy parts of trust to a patchwork of databases, vendors, and state-by-state rules.

An investigation by the New York Times, cited by Fox News, found that in 22 states Uber cleared drivers with serious convictions—child abuse, stalking, assault—so long as the conviction was at least seven years old. Uber has long barred convictions for murder, terrorism, sexual assault, and kidnapping, but its “lookback” logic effectively treated other violent or predatory offenses as time-served reputational debt. Uber defended the seven-year cutoff as balancing safety with “a chance to work and rebuild,” per the Times.

Now, Uber says it will tighten policy: lifetime bans for violent felonies, child/elder abuse and endangerment, and sexual offenses; plus changes around which “less serious” charges still get a seven-year cutoff (harassment, weapons charges), according to the Times via Fox. No timeline has been provided.

The political response is to demand more “screening,” meaning more KYC, more identity verification, and more registry plumbing. But the story’s real vulnerability is not merely whether a background check uses a seven- or ten-year lookback. It’s the definition of “cleared” in a system where identity is a credential, not a person.

Platform safety depends on at least three brittle links: (1) the integrity of the driver’s identity (stolen identities, synthetic IDs, account takeovers); (2) the completeness and timeliness of criminal records across jurisdictions; and (3) the incentives of third-party screening vendors paid to deliver a binary “pass/fail” quickly at scale. Gig-work churn—drivers cycling in and out—makes continuous re-screening expensive, while a single scandal makes it politically mandatory.

Uber’s own safety reporting, cited by the Times, is grim: from 2017 to 2022 the company received a report of sexual assault or sexual misconduct on average every eight minutes. The Times also reported that Uber sometimes allowed drivers to keep driving after rider complaints, only removing them after more serious allegations.

The platform-state symbiosis is visible in miniature. Uber wants the legitimacy of “we screen,” regulators want the comfort of “we required screening,” and both sides can point to process while the actual attack surface—identity fraud, weak account security, and inconsistent records—remains.

When the next incident arrives, expect the usual escalation: stronger KYC mandates, more data sharing, more centralized lists. Those compliance costs are easiest for the largest platforms to absorb—turning “safety” into a barrier to entry, and “public protection” into another moat.

The question isn’t whether Uber should ban violent offenders. It’s whether society wants mobility mediated by a handful of companies whose definition of safety is whatever their vendor can certify—and whose next compliance upgrade will be written by the same political class that can’t tell a stolen account from a reformed human being.