Different people want different things from a relationship: unfailing kindness maybe, or someone tough, with a bit of an edge. Nobody, however, wants a partner who comes with lawsuits, a broken ethical compass, federal agents sniffing around the house, a work ethic that moves beyond crushing to something more sinister, and endless instability.

A partner like Uber. As the ride-hailing outfit barrels into another week of horrifying headlines, with CEO Travis Kalanick taking a leave of absence for an undefined period of time, its short-term prospects appear bleak.* The trouble has long-term implications, too.

If Uber is to thrive in the future, Kalanick has said, it must master self-driving cars. Any competitor that tosses paid human drivers could undercut Uber's service with cheaper rides. But for Uber to get there first—or at all—it needs help. It needs partners.

“The thing about the autonomous car is that it’s an ecosystem that’s going to make it,” says Evan Rawley, who studies competitive strategy and organization at Columbia Business School.

As self-driving tech approaches market readiness, the various players have rushed to join forces. Someone builds the vehicle, someone constructs the sensors, someone writes the software, someone runs the dispatch system that actually picks up the riders. Lyft and Waymo; Waymo and Honda; Autoliv and Volvo; Delphi and BMW and Intel and mobilesye; Lyft and Jaguar Land Rover; Lyft and GM.

But bad news cycles are bad news for collaborations. “I think if you’re a car company or a tech company that needs to partner, and you have multiple choices and one of them doesn’t have the brand value, it might hold you back,” says Karl Brauer, an analyst with Kelley Blue Book.

To win self-driving, Uber will need other people. It has no experience building cars, for example, and no intention of figuring it out. But the folks who do might not need, or want, Uber.

OK, a quick review of what’s gone down at Uber in the past few months—the stuff that will make collaborators itchy. First, the legal trouble: The company is embroiled in a lawsuit filed by Google's Waymo over allegedly stolen self-driving tech, which could lead to a federal investigation and charges. There’s also a federal probe into Greyball, a tool Uber allegedly used to evade local regulators.

Then there are the roiling culture issues. In March, former Uber engineer Susan Fowler accused the company of failing to protect her against sexual harassment and discrimination by coworkers and superiors. That set off a series of investigations—an internal one, which led to more than 20 firings this month, and an external one, by former Attorney General Eric Holder, whose recommendations Uber's board unanimously accepted. Their contents aren’t public yet, but Emil Michael, Kalanick’s deputy and senior vice president for business, left the company Monday.

Combine that with other high-profile departures in the past few months, and Uber’s now on the hunt for a new chief creative officer, chief business officer, chief financial officer, senior vice president of engineering, and chief marketing officer.

Uber's valuation still hovers around $70 billion, but 2017 is taking its toll.

“The GM alliance with Lyft that had looked like a consolation prize for the car company now looks a lot better,” says Erik Gordon, who studies tech entrepreneurship at the University of Michigan’s Ross School of Business. “Uber will be less of a force is in self-driving vehicles than it could have been.” (Uber did not reply to a request for comment on this story.)

That’s too bad, because Uber has what a lot of automakers might need, like a top-notch dispatching algorithm for finding, pooling, and moving passengers, built on billions of rides in hundreds of cities. It’s got reams of data on city traffic. It knows where its riders want to go and when they go there. This is particularly valuable in the testing phase of autonomous vehicle deployment, because Uber has a big, international network in place. A car company can take its hardware, plug in some tech startups’ software, and use the Uber infrastructure to get a fleet up and moving.

Uber does have a couple of deals in place. Volvo has built Uber tech into a fleet of XC90 SUVs for Uber's test fleet. Daimler, Mercedes-Benz's parent company, has signed up to use Uber's software to connect passengers to its own robocars. ("We are in a constant dialogue with our business partners about multiple topics, including compliance & integrity," says Daimler spokesperson Hendrik Sackmann, who declined to comment on Uber in particular. Volvo did not reply to a request for comment.)

But stumbles at this stage in the game could snowball. “Once the self-driving tech is available, if you’re a car company you can essentially re-create Uber’s current capabilities at a much cheaper rate,” says Brauer, the automotive industry analyst. When autonomous vehicles are up and running, all advantages disappear. “When self-driving tech hits at a widespread level, it starts everything over again,” he says.

Kalanick has said this too. "What would happen if we weren't a part of that future? If we weren't part of the autonomy thing? Then the future passes us by," he told Business Insider last year.

Meanwhile, Kalanick has taken a temporary exit from the company. That actually could be good news for Uber’s partnerships. Kalanick has helmed the go-at-it-alone ethos, the brash and sloppy and rude culture that’s gotten the company into trouble—and led to its greatest successes. “If this is a step that could get Travis out of the CEO position, Uber could be better positioned to create partnerships,” says Rawley.

Maybe Uber will figure out how to be a better friend to employees. Then it can start shaking hands with others in the industry too.

*Story updated at 14:10 EST to include news of Kalanick's leave of absence.