TL;DR:
AI background checks and rideshare assault prevention aren’t the same thing. Companies like Uber and Lyft often present them as if they’re the same. They use AI tools to screen drivers and keep an eye on things all the time. These tools can spot some problems. But they also miss a lot, especially when it comes to things that never make it into a police file.
If you ever see a rideshare assault on the news, you’ll notice the companies always respond the same way. They point to their background checks. They talk about their AI-powered safety tools. They mention features in their apps. The message is always clear: We have technology. We care about safety.
Here’s the real problem. The technology is out there. It works up to a point. But AI background checks for rideshare assault prevention have well-known gaps, and the companies know it. They just haven’t fixed them. Let’s talk about those gaps.
What Does Uber’s AI Screening for Drivers Actually Do?
Uber’s AI screening of drivers for safety works in two steps. When someone wants to start driving, Uber runs a background check using outside companies. This check looks for criminal records in thousands of county courthouses across the country. If the person passes, they can drive.
After that, Uber doesn’t just forget about it. Starting in 2018, Uber set up a system that keeps watching for new criminal charges. If a driver is accused or convicted of a crime, the system is supposed to catch it right away. Then, Uber can take away that person’s access to the app without waiting for another yearly check.
On paper, this all sounds pretty strict. But here’s where things get messy. The AI screening of drivers for safety only works with the records it can find. If something never makes it into a criminal file, or if the database misses it, the system won’t catch it.
An investigation from December 2025 found a big problem. In 22 states, drivers with felony convictions could still pass this check if the crime happened more than seven years ago. Sometimes people use fake names. Sometimes states don’t share every record. Sometimes the company’s database just doesn’t cover a certain area.
Uber’s system can only look at what it gets. If it doesn’t get the right info, it can’t do much about it. That’s the weak spot.
Where Does AI Fall Short in Preventing Rideshare Sexual Assault?
Technology to prevent rideshare sexual assault that relies on criminal records can’t spot drivers who’ve never been charged.
Most sexual assaults are never reported. The National Crime Victimization Survey found that roughly two out of three sexual assaults go unreported to the police. So if a driver assaults someone but no one reports it, their record stays clean. AI can watch for new charges. It can’t catch a pattern if there’s never a charge in the first place.
There are warning signs inside the app. Riders leave bad reviews. Some trips get canceled after complaints. Sometimes a driver takes a strange route or makes stops that don’t make sense. All that data is there, and AI can look at it. In fact, Uber made a tool that does exactly this. It’s called S-RAD. This tool tracks 43 different things to score how risky a ride might be, right as it’s happening.
In one court case, documents showed S-RAD flagged a ride as very risky before the driver picked anyone up. The system gave it a high score, 0.81 out of 1. But Uber still sent that driver anyway.
So here’s the twist. The technology to prevent rideshare sexual assault actually worked. The company just didn’t listen to it. That makes a huge difference in court, and it makes a huge difference for riders, too.
What Do Industry Researchers Say About AI Screening Limitations?
Even people who work in background checks say there are problems.
VericorpHR looked at the issue in 2024. They found some big mistakes in AI screening reports. Some drivers were flagged for crimes they never committed. Others got through when the system missed records that should have stopped them. The bottom line from VericorpHR: AI makes things faster, but you still need a person to double-check every step. If you let a computer decide everything by itself, you could end up in legal trouble.
All of this isn’t just true for job checks. It’s the same story with picking rideshare drivers. AI is quicker than a person, but that doesn’t mean it’s better at spotting every risk.
What Technology Could Actually Reduce Rideshare Assault Rates?
There are rideshare driver vetting AI tools out there that could help prevent assaults. Some companies have tried them, but not everywhere, and not for every ride. Knowing what these tools can do matters. It’s about safety, but it’s also about whether companies really tried to keep riders safe.
One big step is using the data the app already collects. The company can watch for strange patterns, like a bunch of low ratings or complaints, or even odd GPS signals. Uber’s S-RAD tool can spot these problems. What matters is what the company does when the tool actually flashes a warning.
Another idea is to check who’s driving in the middle of the trip. Some rideshare driver vetting AI tools can ask drivers to take a photo to prove they’re the same person who signed up. In some countries, this kind of facial check happens while the ride is still going. Uber has tested this, but hasn’t made everyone use it. This helps stop people who use someone else’s account to drive, which is a known risk.
Fingerprint checks are another option. Schools and hospitals use them. They can catch crimes that simple name checks sometimes miss. In California, there’s even a new law being proposed that would force rideshare companies to use better checks and take assault reports more seriously. The fact that lawmakers want to require this shows companies haven’t done it on their own.
When we work on assault cases at the Law Offices of John C. Ye, we look at what safety tools were available and whether the company actually used them. If a company could have flagged a dangerous driver but did nothing, that puts them in a much tougher spot legally.
Frequently Asked Questions
Does Uber use AI for driver background checks?
Yes. Uber uses AI to help with name-based background checks before it lets someone drive. After that, Uber’s system keeps watching for new criminal charges all year long, not just once a year. They also built a tool called S-RAD that rates each ride for risk by looking at driver habits and where the trip goes.
Why do assaults still happen if AI screening exists?
AI only works if there’s already a record in the system. It can’t spot drivers who have never been charged or reported. Most sexual assaults are never reported to police, so a driver’s record can look clean even if something bad happened.
What is S-RAD and why does it matter in court?
S-RAD is Uber’s tool that gives each ride a risk score based on 43 different signs. In a big court case, S-RAD flagged a ride as very risky, but Uber still sent the driver. That choice is a key reason the company faces legal trouble.
Could fingerprint checks stop more rideshare assaults?
Fingerprint checks can catch more crimes than just looking at names. They work even if someone uses a fake name or moves between states. California is working on a new law to make rideshare companies use these tougher checks, since companies haven’t made this standard on their own.
What’s the biggest gap in current AI driver screening?
The main problem is that AI checks only look at criminal records, not at behavior inside the app. A driver with a clean record but lots of complaints or strange driving patterns probably won’t get flagged. The tech to spot these problems exists, but it’s up to companies to pay attention to what it finds.
Can AI screening failures support a negligence claim?
Yes. If the company had AI tools that warned them about a risky driver and did nothing, that can help prove they were careless. The question in court isn’t just whether the tech existed. It’s whether the company actually used it the right way.
Key Takeaways
- Uber’s AI screening starts with background checks before a driver gets approved. There’s also a system that keeps watching for new criminal charges after that. But these tools only find problems that show up in criminal databases.
- AI screening can’t spot drivers who hurt passengers but never got charged. Most sexual assaults never get reported, so someone can have a perfect record and still be dangerous behind the wheel.
- Uber’s own risk tool, S-RAD, once flagged a ride as very risky, scoring it 0.81 out of 1, before the driver was sent out. The technology did its job. The company just didn’t listen to it.
- Industry experts have found even more problems. AI screening can be unfair, get things wrong, and sometimes it’s hard to understand how it made a decision. That’s why you still need a human to check what the computer finds.
- In California, lawmakers are pushing for stricter rules because rideshare companies haven’t done enough on their own. The fact that there’s even a new law in the works says a lot about how much more needs to be done.
The Tools Were There. The Question Is, Did The Company Use Them?
If this happened to you or someone close to you, don’t wait. Reach out to us. Your first meeting is free. It’s private. There’s no pressure at all.
We’ll take a look at what the company actually knew and what their systems flagged. We’ll help you see what your legal choices are.
Ready to talk? Book your free case review.
