The allegations detailed in a new whistleblower lawsuit against a Silicon Valley robotics company read like the first act of a sci-fi suspense movie: a sidelined safety technician plays Cassandra while a robotics company allegedly rushes ahead trying to commercialize a powerful humanoid robot with bone-crunching capabilities. The situation gets more and more sinister—and intolerable for the safety officer—and finally, company leadership allegedly just gets rid of him so they can build their terminators in peace.
These are just allegations, to be abundantly clear, and a spokesperson for the company itself, Figure AI, has told CNBC the safety technician was “terminated for poor performance.” The claims in the lawsuit are “falsehoods that Figure will thoroughly discredit in court,” the spokesperson further claims.
If the lawsuit—framed as a case of alleged retaliatory termination against a whistleblower—is really fiction, it’s the start of a blockbuster. It invokes riveting corporate dramas like Michael Clayton or The Insider, with a dash of Robocop.
You may remember Figure AI. The company released an eye-popping demo of its 01 model last year in which a humanoid robot appeared to respond to spoken, open-ended commands by carrying out tasks of its own choosing. A request for “something to eat” results in the robot gently handing the user an apple, for instance.
The plaintiff, Robert Gruendel, a robotics safety engineer, who once worked in R&D for Amazon according to his LinkedIn, says he only joined Figure after that demo was made. The suit he filed Friday in a federal court for California’s Northern District, claims that in his first week on the job, he discovered that Figure had “no formal safety procedures, incident-reporting systems, or risk-assessment processes for the robots,” and that the only other person responsible for worker safety was an outside contractor with experience in chip manufacturing, not robots.
Most mentions of a robot in the suit concern Figure’s 02 model, depicted below:
Initially, as outlined in the suit, company brass is receptive to these concerns when Gruendel voices them, and CEO Brett Adcock and chief engineer Kyle Edelberg approve a safety “roadmap.” But then, the following ominous conversation with company leadership occurs, the filing alleges:
“Adcock and Edelberg expressed a dislike of written product requirements, which Plaintiff responded to by indicating that their stance was abnormal in the field of machinery safety and of concern to him as Head of Product Safety.”
In the filing, the heads of the company frequently come across as dismissive of the safety officer they themselves hired. The company’s vice president of commercial allegedly says at one point that Gruendel’s safety mandates would be ignored because the CEO “would shoot us if we did it.”
At the start of 2025, the pressure on Gruendel seems to intensify when Adcock, the CEO, supposedly asks Gruendel “what it would take to put Figure robots in the home.” Per the suit, Gruendel, concerned about the robot’s power, and the unpredictability of the AI at its core, designs another “roadmap,” publishes it internally, and holds a meeting about it that the CEO skips. So, allegedly, Gruendel writes a condensed version and sends it to the CEO, but is ignored.
Investors allegedly see a fairly comprehensive safety plan, which they like, after which company leadership downgrades it, an action Gruendel flags to leaders, according to the suit, saying it “could be interpreted as fraudulent.”
Then things get really cinematic in the lead-up to Gruendel’s September 2025 firing. In July, Gruendel conducts safety tests involving just how hard the robot can hit, the suit says. ”During the impact test, [the robot moves] at super-human speed,” and generates force “twenty times higher than the threshold of pain.” According to Gruendel’s calculations, it produces “more than twice the force necessary to fracture an adult human skull.”
The next day, according to the suit, the company’s vice president of growth gets in touch with Gruendel to tell him he had just received a raise in the amount of $10,000 per year with an admiring note about Gruendel’s “continued growth and impact at Figure.” The supposed note also acknowledges Gruendel’s “consistent effort,” and “positive mindset.”
Fresh from receiving his raise, and apparently undeterred, he sends a Slack message to the CEO, saying the robot could inflict “severe permanent injury on humans,” only to be ignored again, the suit alleges. So the suit says he tries the chief engineer, telling him Figure needs to take “immediate action to distance personnel from the robots.”
Gruendel starts worrying, the suit says, that near-misses are occurring, and that there’s no system in place to track them. And then:
”This conclusion was further evidenced by an instance where an employee was standing next to [a robot] and the [robot] malfunctioned and punched a refrigerator, narrowly missing the employee. The robot left a ¼-inch deep gash in the refrigerator’s stainless-steel door.”
So Gruendel, as depicted in the suit, seems to pour everything into getting an emergency stop button added to the robot system in the workplace in order to protect the employees who have to be near it. The company seems to cooperate with the effort, and then more or less abandon it, the suit alleges. Also, a safety feature allegedly gets axed around this time because someone doesn’t like how it looks.
Between mid August and early September, the suit alleges that Gruendel’s authority within the company degrades, and he’s finally fired by the same guy who had praised him and given him a raise earlier that summer.
You can read the whole filing for yourself here.
As CNBC notes, Figure’s valuation has grown 15-fold since last year when it received capital injections from Nvidia, Jeff Bezos, and Microsoft. A funding round this year from Parkway Venture Capital places the company’s value at $39 billion.
As evidenced by the viral reaction to the more recent Neo robot from 1x technologies, there seems to be a race to bring household humanoid robots to market. And there are, of course, bubble concerns accompanying this gold rush-style corporate mindset. In September, roboticist and iRobot founder Rodney Brooks wrote an essay claiming that “today’s humanoid robots will not learn how to be dexterous despite the hundreds of millions, or perhaps many billions of dollars, being donated by VCs and major tech companies to pay for their training.”
Gizmodo reached out to Figure for additional comments about the allegations in this suit, and will update if we hear back.
