Every Eight Minutes: Uber’s Alarming Sexual Violence Problem

34m
For years, Uber has said it is one of the safest ways to travel. But a New York Times investigation found that the company has been contending with a major problem: Hundreds of thousands of people reported that they were sexually assaulted or harassed during Uber rides.

Emily Steel, who broke the story, discusses what executives knew about the problem and how they failed to take certain steps that were supposed to make riders safer.

Listen and follow along

Transcript

you know Tide has been upgraded to provide an even better clean and cold water?

Tide is specifically designed to fight any stain you throw at it, even in cold.

Butter?

Yep.

Chocolate ice cream?

Sure thing.

Barbecue sauce?

Tide's got you covered.

You don't need to use warm water.

Additionally, Tide pods let you confidently fight tough stains with new coldzyme technology.

Just remember, if it's gotta be clean, it's gotta be tide.

Hey daily

And we have a special announcement before today's show, which is that we have made merchandise.

We have made hats, Kevin.

Incredible, beautiful hats.

These hats are available for a limited time as a thank you gift to people who purchase a new annual New York Times audio subscription.

They're black.

They say hard fork on them.

They're very pretty.

And our audio subscription supports shows like ours, as well as shows like The The Daily, The Ezra Klein Show, Modern Love, and Cannonball, the new Wesley Morris show.

So, please, if you haven't yet, go subscribe to New York Times audio and you will get this beautiful hat sent in the mail to your door.

That's right.

So, if you live in the United States, you can subscribe and get your very own hard fork hat at nytimes.com/slash hard fork hat.

It's way more useful than a tote bag.

And plus, you already have a hundred tote bags.

What do you need another tote bag for?

Go get a hat.

From the New York Times, I'm Natalie Kitrowa.

This is the Daily.

For years, Uber has said it's one of the safest ways to travel.

But a Times investigation has found that behind closed doors, the company's been contending with a massive problem.

Hundreds of thousands of people reporting that they'd been sexually assaulted or harassed during Uber rides.

Today, my colleague Emily Steele, on what executives knew about the problem and how they failed to take certain steps that were supposed to make riders safer.

It's Friday, August 8th.

So, Emily, you've covered sexual assault for the New York Times for years.

You've won a Pulitzer for your Me Too coverage, and you've also covered safety issues in a number of industries.

Tell me how you arrived arrived at looking at Uber, because I think this investigation kind of combines your two areas of expertise.

Yeah, so it was back in March.

Someone had suggested that I look into Uber and to look into the company's issues with sexual assault.

And at first I was skeptical.

I thought, you know, Everybody knows about Uber's raucous tech bro period.

Everybody knew about the CEO, Travis Kalanik, who reigned over Uber during this wild time when there was scandal after scandal.

There was a 2014 rape in India that made headlines, and the company came under scrutiny for how it handled that incident.

There were also a lot of issues inside the company where there was sexual harassment within the workplace.

Allegations were very prominent at the company during that time.

And then what happened was the company cleaned up.

Travis Kalanik was pushed out of the company.

Uber hired this new CEO, his name was Dara Kazrishahi.

He was the head of Expedia previously, and he was really seen as the grown-up who had come in to clean up this company's tech bro culture, to get this company ready for its initial public offering.

He introduced a new era, he said, of transparency and integrity and accountability.

And the company promises that they will put safety first.

The motto for the company has switched from super pumped to do the right thing, period.

And publicly, it really seemed like the company had cleaned things up.

And so it made me wonder, is there a story here?

And what did you find?

So then I started to look and I started to dig a little bit and I saw that there were lawsuits against Uber all across the country.

And I saw that there were these two proceedings that were consolidated in California, one in state court and one in federal court, where there were literally hundreds of lawsuits from people who had claimed that they had been assaulted or harassed by their Uber drivers.

And there was a person who said that they were sexually assaulted, battered, raped, harassed, and attacked by their driver.

There was another complaint where somebody said that their Uber driver had thrown this woman's phone and purse out of the vehicle, pulled over and started to do all this terrible stuff to her.

Yikes.

And then I came across a story that was just horrifying.

It happened on a Thursday night in December in 2023.

There was a woman in Houston who called an Uber to come pick her up from an apartment complex.

So she was picked up at 8.53 p.m.

She wanted to be dropped off at this house about 22 minutes away, but the trip diverged

with an eight minute stop near a gas station.

At that point, Uber sent a notification to this woman.

It's an automatic notification to check on her to say, are you okay?

She didn't respond.

And then the ride continued to deviate from its route.

And at 9.29 p.m., it stopped near a Motel 6.

Uber sent another notification to check on this woman.

She didn't respond.

And then four minutes later, Uber tries to contact her with a robocall.

She doesn't pick up.

And this trip remains active on Uber with no recorded movement until the driver marks it complete at 2.01 a.m.

About four hours later, this woman calls Uber

and she's frantic.

She said that she had been intoxicated.

She said that she had been taken to a hotel and raped by her Uber driver.

She said that she woke up in a hotel room with that driver who then fled after she became frantic.

That is just a horrific story.

Yeah, so Uber banned the driver immediately after the incident.

And I was able to get my hands on an internal investigation into what happened.

The driver had received two previous accusations of sexual misconduct against him for making inappropriate comments.

And three automated attempts to try to get in touch with this rider had gone unanswered.

And it says in this report, it says, quote, are our actions or lack of actions defensible?

And what was really scary is that she wasn't alone.

As I started to look more and more into this, I saw case after case after case after case.

And so I thought, gosh, I don't think people know about this.

This is happening in Illinois, in Maryland, in North Carolina, in Colorado, Nevada, Minnesota, California, Florida, just like all across the country.

And the claims were recent.

These weren't claims from 2014.

They weren't claims from 2018.

They were claims from 2023, 2024, 2021, 2022.

It was recent stuff.

And so it made me wonder,

how big is this problem?

What is the scope of this problem?

What does the company know?

What have they done about it?

And what have they not done about it?

How do you go about getting to the bottom of those questions?

That was a really big challenge.

learned initially that there was no central database that collects all of these reports of sexual assault in the rideshare industry.

So I thought, okay, maybe I can build my own.

And I submitted freedom of information requests to cities all across the country to say,

how many police reports, how many complaints of sexual assault do these different municipalities receive?

And that is a long process.

And so I submitted those requests and then kind of kept digging.

And then I talked to a lot of sources.

And what people

said

was that

the place where I could get all of this information would be Uber, that Uber would know how many reports of assault were happening on its platform.

Right.

Uber collects all of that information.

And so I was determined to figure out how I could get my hands on that information.

You got to go to the source.

I had to get the data, yeah.

And it sounds like you did somehow.

I did somehow.

Obviously, I can't say how I got this information, but

over the course of this reporting, I was able to get my hands on

literally thousands of pages of internal company documents.

These are

emails, they are

Slack messages, they are PowerPoint presentations, they are reports and other documents.

Hundreds of these documents have been filed

in court under seal.

Okay, and what do you see when you're looking at these documents?

So one of the first things that catches my eye when I'm looking at these documents is a

spreadsheet.

And it says Uber trips with reports of sexual assault or sexual misconduct in the U.S.

And what it says is that Uber received a report of sexual assault or sexual misconduct in the United States almost every eight minutes on average between 2017 and 2022.

One report to Uber of sexual misconduct or assault every eight minutes?

What does that amount to?

So the total number for that period is 400,181.

Wow.

That is much more than what the company has publicly disclosed.

Publicly, the company has said that they had 12,522

accounts of serious sexual assault on the platform during the same time period.

But Uber had not revealed the total number of sexual assault and sexual misconduct reports that it had received.

Okay, so this is a difference basically between what Uber is saying is serious sexual assault and this bigger, far bigger number that includes all sexual assaults and claims of harassment.

Exactly.

Uber tracks 21 different categories of sexual assault and sexual misconduct.

What it reveals publicly is five of what it considers the most serious categories of sexual assault.

And this number, this 400,000 figure, represents all of the cases of sexual assault and sexual misconduct on the platform.

Did you ask Uber about this?

What did they say?

I did ask Uber about this.

And what Uber has said is that there is no tolerable level of sexual assault.

And they did point out that about 75% of those 400,000 reports were quote-unquote less serious in nature.

They were things like making comments about somebody's appearance, flirting, or using explicit language.

The company also said that those 400,000 reports had not been audited by the company, so they could include incorrect reports or also fraudulent reports that were submitted by people who wanted a refund.

Okay, so they're saying 75% of that 400,000 is less serious.

But I guess, I mean, to me, you know, 25% of that amount is still like 100,000 claims.

That's a big number.

What Uber says is millions of people hail a car in the U.S.

every single day, and that the vast majority of those trips go without incident.

They say 99.9% of rides occur without any report, any safety report of any kind.

And they say that because Uber operates on such a large scale, that you're going to see the best and worst of society show up on those rides.

But the thing is that even a fraction of a percent can translate into many, many attacks.

So it sounds like you're seeing the answer to your first question that you came to these documents with, which is what's the scope of this problem?

And that answer is, it is big.

It's a big problem.

What did you learn about your second question, which is, what did Uber know?

And when did they know it?

What did they do?

So as I'm paging through these documents, it's literally thousands and thousands and thousands of pages.

And what I found was that Uber not only was aware of this problem, but they actually were studying incidents of sexual violence by 2016.

And that year, there were data scientists in the company, and they saw that sexual assault followed distinct patterns on the platform.

What do you mean by that?

What are the patterns?

So women most often are the victims, whether they are passengers or drivers.

The attacks typically occur late at night and on the weekend with pickups originating near a bar.

In the vast majority of these cases, the offenders are men, either drivers or passengers.

And those men, both the drivers and passengers, have records of sexual misconduct complaints on Uber and low ratings.

And then intoxicated passengers are especially vulnerable.

It's interesting.

We're not just talking about female passengers, right?

It's also female drivers who are experiencing harassment and assault at the hands of passengers.

Right.

And that was something something that wasn't an issue that I immediately knew about, but it was something that I learned about in this reporting: that drivers are victims too.

It's not just passengers.

And one of the other really interesting things that I found is that for years,

teams inside of Uber have tested tools that they found were effective at making trips safer,

but the company delayed or did not require its drivers to adopt some of those most promising programs.

And they also didn't warn passengers about the factors that it linked to attacks.

We'll be right back.

KPMG makes the difference by creating value, like developing strategic insights that help drive M ⁇ A success and embedding AI solutions into your business to sustain competitive advantage or deploying tech-enabled audits to deliver more accurate and transparent outcomes, brighter insights, bolder solutions, better outcomes.

It's how KPMG makes the difference every day.

KPMG, make the difference.

When you need a break, skip to scrolling.

Visit myprize.us.

The games are super exciting and you can actually win.

Myprize.us is the most fun, free-to-play social casino around.

Everyone deserves to win big.

All the slots and table games you love with incredible bonuses.

Sign up today for an incredible welcome package.

Myprize.us is a free-to-play social casino.

Users must be 18 or older to play.

Voidwear prohibited by law.

Visit myprize.us for more details.

When you need a break, make it memorable.

Visit myprize.us.

Real prizes, real winners, real easy.

My name is Diana Dunn.

I would like to be able to share my New York Times account with my granddaughters.

We live in a very rural area.

I would like to share my account just to increase their knowledge of our whole world.

Thank you so much.

Diana, we heard you.

Introducing the New York Times Family Subscription.

One subscription, up to four separate separate logins for anyone in your life.

Find out more at nytimes.com slash family.

Emily, let's talk about these tools that you said Uber developed that the company itself thought could kind of help prevent sexual assault during rides.

Tell me about those solutions.

Yeah, so I think there are three worth talking about.

One was designed to be part of the sophisticated algorithm that happens behind the scenes when you open up the Uber app and tap to request a ride.

There were teams of data scientists inside of Uber who for years

worked to develop this special model to forecast which matches between a driver and a passenger might lead to a sexual assault.

How do they do that?

So there was this one presentation that I looked at.

It was a 2017 presentation.

It was called Preventing Sexual Assault.

And this document detailed how the company could use 43 different predictors, things like reporting a creepy driver, safety incident reports, geographic information, like the number of bars near a pickup.

And this data-driven machine learning model does some sort of sophisticated calculation to score the risks of pairings between drivers and passengers.

And the company actually quietly tested this program in Los Angeles in 2018.

And what Uber found is that this program could correctly anticipate 15%

of sexual assaults on trips that were using Uber's basic ride hailing option.

Sorry, so are you saying that Uber was running a safety test on like live customers where the company is accurately predicting ahead of time that someone is likely to be assaulted or harassed?

That's what these documents show.

They show that the company was testing this model in shadow mode and not acting on it.

An internal presentation a few months later on this tool called it potentially the most effective intervention for preventing sexual assaults.

So what do they do with it?

So what I learned is that this tool has been up and running in the United States since 2022.

So it's part of that system that Uber now uses to respond when you request a trip.

And it has long been considered a trade secret at the company.

And the company hasn't spoken publicly about this technology until now.

The way that this tool works, it first determines the risk of potential pairings between drivers and passengers.

and then it uses those scores to try to select the best match.

And so it will prioritize rides that are safer and downrank pairings between drivers and passengers that are less safe.

So the rides that are determined as risky are deprioritized.

They're put at the end of the line of all the options that you might get when you call an Uber, right?

Exactly.

But what I found in all of these records was a document from 2024 and it identified a problem with this system.

It said

that Uber was still dispatching trips that it had identified as high risk.

So why not just eliminate those pairings?

Why not, for example, kick off those drivers or riders that are seen as potentially, you know, aggressive so Uber does say that it bans drivers and passengers that are accused of serious sexual assault or a number of lower-level complaints.

And what it also says is that this technology cannot reliably predict whether an individual pairing or trip will result in an incident or not.

And Uber also said that if they were to unilaterally block all sorts of trips, like requests from bars late at night, that that could leave many people stranded on the street and it could encourage them to drive drunk or walk home unsafely.

So they're making the argument that, look, this tool might be useful, but it isn't completely predictive.

And trying to mitigate the risk on any one pairing could actually lead to more dangerous scenarios.

Right.

So what's the second solution that the company looked at?

So another solution that the company has been exploring for years is the idea that you could pair women drivers with women passengers.

Theoretically the people least likely to assault each other.

Right.

The company first introduced this option in Saudi Arabia in 2019 with the choice for women drivers to select women passengers.

But Uber held off introducing this as an option in the United States.

How come?

Well, so what's really interesting though is as recently as last fall, executives green lit a pilot for a women matching option in the United States.

And it was supposed to start in November.

And then days after Donald Trump is elected president, executives decided to hold off.

And there was an internal document that said, this is not the right environment.

to launch and we want to take a beat to reassess our timing.

Hmm.

I'm just wondering, overall, is this really a long-term solution to the problem?

Because there aren't that many women Uber drivers.

At least in my experience, there's way more men driving for the platform.

So one of the big issues or one of the big hurdles to introducing this option is Uber had to have enough women drivers.

Sure.

And there was an internal document that I found that said that several high-profile assaults had made it difficult to attract those women to drive for Uber.

It said, safety has ranked as one of the top reasons for prospective female drivers to not join Uber and for current female drivers to not drive during certain hours of the day.

There also are other things to think about.

People inside the company worried, if we introduce this, are we telegraphing that other Uber rides are not safe?

There were worries that if this option were to cost more, would that be considered a pink tax?

And most importantly, there was the fear of gender discrimination and other lawsuits that the company calculated could cost more than $100 million.

But despite all of that, the company did announce at the end of July that it was going to start testing women matching options in Los Angeles, San Francisco, and Detroit.

It sounds like in the case of both of these first two options, you have them rolled out in one case without us knowing.

In the other case, there was a long delay before it happened.

What about the third solution?

So the third solution is dash cams.

As early as 2014, the company had been exploring the idea of using cameras and cars.

And then the new CEO comes in, Dara Kajwashahi, and he outlines this vision of installing inexpensive cameras in Uber cars that would record trips by default.

Employees studied this plan and they found not only would it be feasible and cost-effective, but it also was likely to reduce sexual misconduct and help drivers.

But the company didn't do it.

Why not?

Largely because installing cameras in Uber cars would conflict with the company's business model.

How so?

What do you mean by that?

So by design, Uber drivers are classified as independent contractors, not employees.

That means that Uber and other gig economy companies can avoid paying significant costs that are associated with traditional employment.

So that's things like minimum wage, it's overtime, it's employee benefits, it's time for rest breaks.

This is the key to Uber's business model.

And Uber goes to great lengths not to treat drivers like employees.

And so they do things like limit trainings and not require certain equipment like cameras to be installed in cars.

This seems pretty critical to the whole story.

I mean, preserving that contracting model is obviously so important to Uber.

You can imagine how much their costs would go up if suddenly they had all of these drivers as employees.

Right, it's crucial to the company's business.

And what Uber says is that it offers the option through its apps for drivers and passengers to record audio and for drivers to use their phones and the front-facing camera on their phone to record video.

But that the decision not to require video recording had as much to do with very practical challenges and privacy concerns as it did with driver's employment classification.

So Uber has essentially come up with a hybrid solution.

And the advantage for Uber is that those options preserve the company's business model.

Just to explore for a second what I imagine that the company's position is.

If the drivers are not employees of Uber, Is it fair actually to hold the company responsible for what happens in their cars?

You know, it's like, I think we all know that when we order an Uber, we're getting into a private car that's being driven by a gig driver.

Sure, they've been vetted to some extent by the company, but Uber is just the matchmaker, right?

They aren't literally providing the service.

You are right.

That's the company's position.

And it is at the core of the litigation that Uber is now facing in these 3,000 lawsuits in federal and state courts from passengers who said that they were sexually assaulted or harassed by their Uber drivers, the company has called those lawsuits a legally baseless attempt to hold it liable for the extremely rare and unforeseeable criminal acts of independent drivers.

In some cases, the company has even turned around and sued those drivers that were accused of assault, essentially saying that if those allegations were true, that the drivers had violated their agreements with the company.

But the people suing Uber claim that the company is responsible for the misconduct of their drivers.

They say that the company is more interested in growth than protecting its passengers.

And then they say that the company has gone to great lengths to cover up what they've described as a sexual assault epidemic.

In just thinking about the three solutions that we've talked about, it does seem as though in each case, there was an aspect of this that the company wasn't really publicizing either the problem itself or the solution in some cases.

And I guess I'm wondering why you think that is, because it does seem like something that, you know, writers would have wanted to know about.

Right.

So in my reporting, that was something that was really striking to me because at the same time that Uber was studying this problem, developing tools for this problem, they also were deploying some pretty aggressive media campaigns to publicize that Uber is one of the safest options for travel.

And to highlight these risks could present a problem.

It could say,

hey,

we might not be as safe as what we said we were.

So you're saying basically one of the reasons the company might not be telling people about this is that it could freak people out.

It could scare them away.

It could contradict that message of safety.

And the idea of warning customers about the patterns linked to attacks has actually caused some rifts inside of the company.

There have been people who have said internally, look, we need to tell people, we need to tell them the rides that are the most risky are the ones when people are getting picked up by a bar.

If it's a single woman in the back seat alone, if she's intoxicated.

And the company ultimately decided not to.

I found a presentation discussing whether or not the company should include day or time patterns about when these incidents occur.

And there's a slide in this presentation that basically says, if we include this information, it will lead to fewer people riding in Ubers.

Huh.

So directly saying the quiet part out loud, this could hurt our business.

Right.

It does seem as though a lot of this comes down to transparency.

I mean, whether Uber is liable is sort of for the courts to decide.

But the fact that the company is really not making this public, isn't being honest with us about what the risks are, and then letting us decide if we want to get in the car.

I mean, that feels like a really important question in all of this.

Aaron Ross Powell, and that's something that came up over and over again in in my reporting is how much does the company know?

How much has it told us about what it knows?

And what should Uber tell us about what it knows?

And if people know about the harm and they know about the risks, they can take steps to protect themselves and to protect others.

Right.

So maybe they won't take an Uber by themselves when they're drunk coming home from a bar.

Sure.

Maybe they won't put their drunk friend in the back seat of a car.

Maybe they will will turn on the options to record the audio of those rides.

You see the stats and you see the data and you see how there were 6.3 billion rides that occurred during this time period when all of these assaults occurred.

That's just a fraction of a percent of rides that end in serious incidents.

But when you look at those individual incidents, it's stories of people

whose lives have been upended by one ride

and

they're facing a lot of trauma and a lot of pain.

And

the question is,

if they had had more information, would they have been better able to protect themselves?

And did the company do enough to protect its customers?

Emily, thanks so much for your time.

Thank you.

We'll be right back.

It's your headline to unpack.

It's your one story to follow week by week.

It's your Wordle to work through.

It's your team to track.

It's your 36 hours to explore.

It's your marinade to master.

It's your opinion to figure out.

It's your mattress to upgrade.

It's your day to know what else you need to know today.

The New York Times.

It's your world to understand.

Find out more at nytimes.com/slash your world.

Here's what else you need to know today.

After 10 hours of deliberations, Israel's leaders approved Prime Minister Benjamin Netanyahu's proposal to expand the war by taking control of Gaza City, according to an announcement by the government.

The move goes against the advice of the Israeli military.

But the announcement appeared to stop short of saying that Israel would take full control of the Gaza Strip, something Netanyahu said earlier was his goal.

In an interview on Fox News on Thursday, before the Security Cabinet met, Netanyahu said Israel did not want to maintain permanent control of Gaza or govern it.

The Security Cabinet also approved several principles for ending the war, including the return of all remaining hostages, the disarming of Hamas, and the establishment of an alternative governing body in Gaza that doesn't include Hamas or the Palestinian Authority.

It'll likely take the military days, if not weeks, to call up reserve forces and to allow time for the forced evacuation of tens of thousands of Palestinians from the new areas of combat.

And

President Trump has announced he's nominating Stephen Myron, his top economic advisor, to serve as a governor at the Federal Reserve.

Myron has been a staunch defender of Trump's economic policies and has echoed the president's criticism of Fed Chair Jerome Powell for not lowering interest rates.

Trump made it clear that Myron would serve in the position temporarily if confirmed by the Senate, though it's possible that he could stay on.

The White House is planning to run a separate process to nominate someone to fill the seat starting in February, and that person could potentially be elevated to become the chair of the central bank next year when Powell's term expires.

Today's episode was produced by Olivia Knatt, Jessica Chung, and Eric Krupke.

It was edited by Maria Byrne and Michael Benoit,

contains original music by Marion Lozano and Dan Powell, and was engineered by Alyssa Moxley.

Our theme music is by Jim Brunberg and Ben Landsberg of Wonderly.

That's it for the daily.

I'm Natalie Kitroweth.

See you on Monday.

Gun injuries are the leading cause of death for children and teens in the United States.

Some people avoid talking about gun violence because they don't think they can make a difference, but every conversation matters.

When it comes to gun violence, we agree on more than we think, and having productive conversations about gun violence can help protect children and teens.

Learn how to have the conversation at agreetoagree.org, brought to you by the Ad Council.