On a Texas evening last week, a 76-year-old grandmother named Martha Avila was standing in the front room of her suburban home when a Tesla Model 3 hurtled into her brick home at a reported speed of over 70 miles per hour, killing her.

The car’s driver, 44-year-old Michael Butler, later told police that he had Tesla’s driver assistance features—which the automaker argues make driving safer and less stressful—engaged during the crash. Butler exhibited “no signs of intoxication,” the Harris County Sheriff’s Office, which responded to the crash, noted in a report.

Now Avila’s family is suing not only Butler but also Tesla, alleging that the electric-auto maker’s Full Self-Driving (Supervised) driver assistance feature, also called FSD, played a role in her death. The feature is designed to handle certain aspects of driving—including navigating city and residential roads, stopping for red lights and stop signs, and changing lanes—but it requires drivers to pay attention and stay ready to intervene if the system makes a mistake. The suit alleges Tesla’s tech “was defective in design and unreasonably dangerous,” lawyers representing Avila’s daughter and son-in-law wrote in a lawsuit filed in Harris County District Court on Tuesday. (The son-in-law, Justin Barbour, was also in the home and injured in the crash.)

Tesla didn’t respond to WIRED’s request for comment. But on X, Tesla vice president of AI software Ashok Elluswamy wrote that Tesla data showed that Butler “manually overrode self-driving by pressing the accelerator all the way to 100 percent” and “had the accelerator pressed even after the crash.” Tesla CEO Elon Musk posted that speculation that the company’s technology played a role in the crash “makes no sense.”

Plenty of the crash’s specifics have yet to come out, and it’s very possible that Tesla’s tech didn’t have anything to do with Avila’s death. But even if the driver is mostly responsible for what happened, the electric-auto maker could still be found at least partially culpable—and liable for big monetary damages.

“If the product is designed in a way that it leaves drivers vulnerable to situations where suddenly the system is not working and they’ve lost situational awareness, Tesla could be found responsible,” says Matthew Wansley, a professor with Yeshiva University’s Cardozo School of Law who studies automotive tech.

In fact, it’s happened before. Last year, a Florida jury found that the driver of a Tesla Model S using Autopilot, Tesla’s earlier driver assistance software, was mostly responsible for a crash in which he failed to see that the T-shaped intersection his car was traveling on was ending. He kept his foot on the accelerator, and the Tesla collided with and killed 22-year-old Naibel Benavides Leon. Her boyfriend, 26-year-old Dillon Angulo, was seriously injured. (Despite often touting its vehicles’ expansive data collection abilities, Tesla said it couldn’t recover critical data related to the case; Benavides’ family lawyers were later able to recover it with help from a hacker.)

But the jury also found, in a precedent-breaking decision, that Tesla shared one-third responsibility for the crash because it believed Autopilot was effective. It determined that Tesla was liable for $200 million in punitive damages, plus an additional $43 million in compensatory damages. A judge upheld the verdict earlier this year.

Critics of Tesla’s approach argue that it’s precisely because FSD is pretty great that the feature presents a problem. If drivers trust that the system operates well all the time, they might not be prepared to take over if something goes wrong. In a 2018 California highway crash, the driver behind the wheel of a Model X using Autopilot failed to take over steering before the vehicle crashed into a barrier, killing him. (Tesla later settled a lawsuit related to the crash hours before it was set to begin.)