SPECIAL REPORT-Tesla is browbeating the EU to approve its ‘Full Self-Driving’ tech. The pressure is working

By Chris Kirkham, Abhirup Roy and Marie Mannes Oct 7 - As Tesla chases an elusive prize – government approval of its “Full Self-Driving” technology across Europe – the US automaker is waging a pressure campaign casting safety regulators as feckless bureaucrats blocking its life-saving technology.

Elon Musk's company has pushed regulators directly to reduce the rigor of safety reviews, submitted its own research making bold crash-reduction claims, and used social media to mobilize Tesla enthusiasts to demand speedy approvals. The ​strategy has worked so far, winning a critical first approval in the Netherlands that cleared the way for a Europe-wide vote expected soon by regulators from 27 European Union countries. Seven other EU countries, including the Czech Republic, Belgium and Denmark, have granted Tesla national approvals for the driver-assistance technology, which is not fully autonomous.

But Tesla’s campaign is built on faulty company safety research that can’t support the company’s claims to officials or the public, according to seven traffic-safety researchers ​who examined its methodology for Reuters. The automaker sent the flawed studies broadly to European regulators seeking their support after it secured approval in the Netherlands. The Dutch evaluation had outsized importance because the EU’s process required first getting signoff from one member country that agrees to endorse the technology for EU-wide approval. Other ‌countries’ regulators are now relying heavily on the Dutch ​research to decide their own votes.

During that Netherlands review, Tesla repeatedly sought to scale back the intensity of the regulator’s examination, and it pushed – with some success – to set the terms and methods of its evaluation and testing, according to a Reuters review of correspondence between the automaker and the regulator, which has not been previously reported. Dutch staffers often responded with conciliatory notes emphasizing the regulator’s desire to collaborate with Tesla on “shared” or “mutual” objectives, raising concerns about its impartiality among five autonomous-vehicle regulation experts who reviewed the exchanges for Reuters.

Tesla did not respond to detailed questions from Reuters on its safety research and interactions with European regulators. RDW, the regulator responsible for vehicle safety in the Netherlands, didn’t comment on its correspondence with Tesla or the company’s European data. In a statement, RDW said it “assesses driver assistance systems from all manufacturers objectively and independently.” It added that engagement with a manufacturer “is a routine part of the assessment process and does not compromise the independence or thoroughness of our assessment.” For this report on Tesla’s European pressure campaign, Reuters obtained hundreds of pages of email correspondence between Tesla and regulators in the Netherlands, Sweden, Norway, Denmark, Estonia, Italy and Switzerland; reviewed safety studies Tesla distributed to regulators across Europe; and interviewed seven traffic-safety researchers who examined Tesla’s crash-reduction claims. Together, the documents and interviews provide the first detailed account of how Tesla sought to influence the Dutch review of FSD and used flawed claims about the technology’s life-saving potential to press regulators across Europe.

Tesla produced an “evidence dashboard” on Full Self-Driving (FSD) safety that it said it shared with EU ‌member-state regulators in April. The company’s Europe-specific research compares FSD-equipped Tesla cars driven by its own professional engineering drivers to all Tesla cars driven by ordinary owners across Europe. But it couldn’t determine relative crash rates because its sample of miles driven in FSD-piloted vehicles was too small – in the tens or hundreds of thousands – to meaningfully compare to the hundreds of millions of miles covered by manually driven Teslas in Europe, according to the seven traffic-safety researchers’ review of company research. Instead, Tesla used a set of driving behaviors as “surrogates” for measuring crashes – horn honking, low blinker-usage frequency, rapid acceleration or deceleration, automatic emergency braking and hard cornering. The company says it can track such maneuvers through its in-car technology, but it hasn’t sought outside peer review of its data or its methodology for analyzing it. The traffic safety researchers told Reuters that some of these driving behaviors, such as honking and blinker usage, are poor indicators of crash danger. Others, like hard braking, can be better proxies. But they have their own limitations, the researchers said, and their relation to crash rates isn’t well-established.

All of the researchers said Tesla’s study provides no evidence of the central claim in the company’s European lobbying campaign – that FSD prevents fatal crashes. The researchers also questioned Tesla’s use of professional engineering drivers as a comparison. Trained drivers are more familiar with how FSD works and when and how they might need to intervene to prevent crashes or avoid the specific driving behaviors Tesla tracked. Tesla calls the system “Full Self-Driving” but adds “(Supervised)” because it requires a human driver's full attention. The automaker warns drivers to pay strict attention and stay ready to take over if FSD fails – transitions that autonomous-driving experts say can be difficult and dangerous.

In its Europe study, Tesla says the professional drivers’ intimate familiarity with FSD makes its data collection more accurate because they rarely turn the system off. In the Europe study’s fine print, Tesla warns that its comparisons of FSD driving and human driving behaviors can’t be interpreted as “causal estimates,” an acknowledgment that FSD’s performance on those measures won’t necessarily translate to lower crash rates.

But Tesla and Musk don’t disclose the research’s limitations in their social-media posts blaming European regulators for traffic deaths. When French transportation minister ‌Philippe Tabarot said in July that his country couldn’t support EU-wide expansion of FSD due to safety concerns, Musk posted on his social media platform X that “delaying the approval of FSD in France will cost lives.” Such claims go far beyond what Tesla’s research can support, said Marco Benedetti, an assistant research scientist at the University of Michigan Transportation Research Institute and a former statistician at the US National Highway Traffic Safety Administration (NHTSA).

“When you start getting into ‘lives saved,’ that’s where it gets nonsensical,” he said. TESLA’S LONG HISTORY OF EXAGGERATED SAFETY CLAIMS

Tesla’s inflated safety claims to European regulators are the latest in a long history of company exaggerations involving FSD’s safety and readiness for fully autonomous driving. The automaker has for years faced US federal investigations and lawsuits over crashes and deaths blamed on its FSD or Autopilot driver-assistance systems. Every year for a decade, chief executive Musk has asserted Tesla would deliver fully autonomous vehicles no later than the following year. After launching a small robotaxi service in Austin, Texas, in June 2025, ‌Musk promised “hyperexponential” growth of a technology Tesla says can work anywhere on earth. Instead, Tesla has expanded robotaxis slowly in Austin and a handful of areas in Texas and Florida.

The bulk of Tesla’s $1.5 trillion stock-market value is tied to investors’ belief that it will dominate self-driving technology globally. A Reuters investigation in May found the automaker is nowhere close to deploying self-driving cars safely at scale and that it used unsound statistical methodology to claim FSD is already up to 10 times safer than a human driver. The automaker, for instance, relied on an apples-and-oranges comparison of crashes in FSD-piloted Teslas with all crashes nationally. For Teslas using FSD, the company counted only crashes with airbag deployments; it then compared that rate with all crashes in federal data where a tow truck removed a vehicle – a criterion that includes a far higher number of less serious crashes.

Tesla submitted the same misleading US research to European regulators, along with its more recent Europe-specific study. Some of the traffic-safety researchers interviewed by Reuters said Tesla’s European study marked an improvement over its US research and included more disclosures acknowledging shortcomings of its methodology. Further, they said, making safety comparisons of a system like FSD to a human driver are inherently difficult. That’s in part because human drivers still play a critical role in operating FSD safely, making it difficult to attribute any perceived benefit solely to FSD. Human drivers can also turn the system on and off, and tend to avoid using it in complex driving situations they perceive to be dangerous, the researchers said.

The main problem, all of the researchers said, is that Tesla makes exaggerated public safety claims that go far beyond what its research proves in both the US and Europe. Daniel McGehee, an engineering professor and director of the University of Iowa's Driving Safety Research Institute, said Tesla’s European study presents "potentially useful evidence of system performance" but stops well short of showing FSD saves lives. FSD approval is critical to Tesla’s ambitions in Europe, where the company is only now regaining market share after steep sales declines stemming from rising electric-vehicle competition and consumer backlash over Musk’s backing of US President Donald Trump and far-right European politicians.

Tesla now competes with dozens of new EVs from European brands and from surging Chinese EV makers, many of which offer FSD-like driver-assistance technology in their home market. FSD is currently available in the United States, where it didn’t need regulatory approval, along with the European countries that recently approved it and a handful of others, including Canada, Mexico, Australia and South Korea. The automaker has struggled for years to secure approval in China, its second-biggest market behind the US.

Tesla would be the first ⁠automaker to win EU approval for an advanced driver-assistance system allowing ​hands-free operation on any road. Currently, regulations only allow hands-free systems on highways. In the EU-wide vote, Tesla needs support from regulators in 55% of EU-member countries constituting 65% of the region’s population. REGULATOR’S ‘SHARED GOALS AND OBJECTIVES’ WITH TESLA

Staffers of the Dutch regulator, RDW, eased aspects of its FSD safety ⁠evaluation in response to repeated complaints from Tesla as it reviewed the system starting in 2024, correspondence between the company and the regulator shows. RDW ultimately found FSD could improve safety if used correctly and recommended EU-wide approval. The exchanges, which have not been previously reported, raise concerns over whether Tesla exerted undue influence over the safety review of its own technology, according to autonomous-vehicle regulation experts who reviewed the correspondence for Reuters. Throughout the process, Tesla representatives asserted the company should have shared decision-making power in how the review would be conducted. Staffers for the regulator, in turn, often assured Tesla of their desire to work collaboratively toward mutual goals.

In November 2024, early in Tesla's discussions with RDW, the company pushed the regulator to start testing vehicles by month’s end. When a regulator responded with a commitment to deliver Tesla a project plan by then, a responding Tesla staffer reiterated the demand for immediate testing. “Keep in mind that this is mission critical for our leadership,” the Tesla representative wrote. “Our leadership will measure the goodwill to do this project with RDW by this ‘test start’ milestone.”

The start of testing took longer. But Tesla won concessions from Dutch regulators after complaining about the costs, methods and time needed for other aspects of the review, according to the exchanges between the company ⁠and RDW. Unlike in the US, automakers in Europe must pay regulators’ expenses for evaluation of new-vehicle models, systems or parts. The costs disputed by Tesla ranged from tens to hundreds of thousands of euros, small sums relative to the $95 billion in revenue posted last year by the world’s most valuable automaker.

Tesla objected, for instance, to paying for five RDW staffers to travel to a May 2025 meeting at Tesla’s Gigafactory in Berlin. The company said it expected only two to three staffers and argued most conversations should be held remotely. RDW responded that only two staff members' expenses would be billed to Tesla and others might not attend. In an April 2025 email to RDW, a Tesla employee balked at RDW’s breakdown of €365,000 in costs for testing permits, technical work and other services RDW conducted over a five-month period ending in March 2025. The automaker complained of the “extremely high cost” and said it would withhold payment until it received a "granular" accounting of RDW's operations, including credentials, hours worked and hourly cost of every RDW employee assigned to the project.

An RDW staffer offered to dial back the review’s intensity a week later, inviting Tesla to “share any suggestions on how we might adjust or reduce the ​required effort moving forward.” RDW redacted all the names of Tesla and RDW staffers, along with large portions of their correspondence, before providing the documents to Reuters. The extent of the regulator’s redactions, where entire emails or large parts of exchanges are obscured, leaves substantial gaps in the dealings between Tesla and RDW.

Bryant Walker Smith, a University of South Carolina law professor focused on autonomous-driving regulation, questioned whether Tesla was challenging relatively inconsequential expenses as a way to reduce regulatory scrutiny. "These don't seem like huge amounts,” he said. "My immediate concern was: Is reducing costs a proxy for reducing oversight?"

In December 2024, Tesla complained that travel logistics for an RDW inspector to “witness” certain tests were wasting valuable time and that test reports submitted by Tesla should suffice. RDW responded by proposing to monitor part of the tests remotely. The type of test being performed was unclear in the documents because portions of them were redacted by RDW. In a January ⁠2025 email, Tesla questioned RDW’s request for a meeting to discuss Tesla’s remote software system, or over-the-air (OTA) updates, and instead offered a written summary. The automaker and the regulator could "decide together if an OTA scrutiny meeting is justified,” a Tesla employee wrote.

An RDW staffer, in a partially redacted email, assured Tesla that the regulator didn’t want to “dive too deep in technical details,” and didn’t “question the quality and process for OTA’s.” In an April 2025 email to RDW, Tesla said some tests that the regulator had conducted in Lelystad, Netherlands, “should have been handled in a different way” and demanded that RDW get written approval before certain tests of FSD on its vehicles. Tesla also told RDW it needed to tell the company in advance exactly what kinds of tests it would conduct and provide a list of names and roles of its safety drivers and passengers performing them.

RDW responded two days later and agreed to provide the names, after getting sign-off from the drivers, and highlighted its "shared goals and objectives" with Tesla. An RDW staffer wrote: "We hope you recognize that our efforts are driven by the best intentions to support and achieve those mutual aims." The regulator’s redactions appear to obscure the type of testing it was discussing with Tesla. RDW has shared data on its FSD testing with other European regulators but declined to release it publicly. Both Tesla and RDW argue such information would ⁠reveal Tesla trade ​secrets, the correspondence between them shows. Some transportation regulation experts disagree.

RDW's accommodation of some Tesla demands conflicts with the role of a regulatory watchdog seeking to protect public safety, said Nynke Vellinga, a senior researcher at University of Groningen in the Netherlands, who specializes in legal aspects of autonomous driving. "An authority like the RDW needs to be very strong in saying: No, this is what we deem necessary,” she said. “And if we cannot do it, that means that you're not getting approval."

The regulator RDW told Reuters that it conducted an “extensive independent investigation,” which included testing on closed tracks and public roads, in various conditions. RDW declined to specify how it measured safety performance during such testing. Five of the regulators from the seven other EU countries that have approved FSD so far – Denmark, Belgium, Croatia, Estonia and Slovenia – declined to comment or did not respond to questions about Tesla’s safety research, its social-media pressure campaign and exchanges between Tesla and RDW provided by Reuters.

Officials in Lithuania and the Czech Republic said the company’s social media claims did not influence their assessment of its technology. The Czech regulator added that it has no reason to question RDW’s impartiality. RIDICULING REGULATORS, FIRING UP THE FAN BASE

At Tesla’s annual shareholder meeting last November, Musk mocked European regulators, describing “crazy conversations” with them that he said were “like a Franz Kafka novel.” “People’s lives are at stake here,” he said. “So definitely, pressure from our customers in Europe to push the regulators to approve would be appreciated.”

Boasts about life-saving technology were a key feature of the misleading US safety research that Musk and other Tesla leaders regularly touted last year as the company’s board successfully sold investors on approving a compensation package awarding Musk up to $1 trillion in Tesla stock. The company leaders regularly cited the claim in its US research that FSD-piloted Teslas are 10 times safer than the typical US vehicle. The company later revised the research and downgraded the claim to seven times safer.

The company created a US-focused safety website saying FSD could save more than 32,000 lives and prevent more than 1.9 million injuries annually. Traffic-safety researchers who reviewed Tesla’s safety research and statistics for ⁠Reuters called those figures essentially meaningless. That’s because they were based on the unrealistic assumption that every US vehicle would be replaced by an FSD-enabled Tesla – and that every vehicle was, in fact, far safer than the one it replaced, as Tesla inaccurately contended. Tesla has made similarly misleading claims in Europe, mischaracterizing its own research, according to experts who reviewed its methodology. Last month, Tesla said on X it had decided to “open-source” the European safety study it said it had given months earlier to European regulators. Tesla called its study “one of the key pieces of evidence used to support the Netherlands approval.”

In posting the safety research, which doesn’t address fatalities, Tesla said that an average of 53 road deaths per day in Europe are “largely preventable with technology we have today.” “Yet in much of Europe,” Tesla said, “that safety benefit remains locked behind bureaucracy.”

Tesla supporters across Europe have used the same rhetoric in angry posts on X blaming regulators for traffic deaths. After Musk accused French transportation minister Tabarot ⁠of endangering lives in July, Tesla enthusiasts pounced on Tabarot’s X account. One user said that FSD would have kept him safe recently as he was tired driving home from work and drifting in his lane. “YOU put me – and other road users along my route – in danger!” the post said, in French. “From ⁠now on, YOU are RESPONSIBLE for accidents and deaths on the road.”

Another user, replying to Tabarot’s account, wrote: “Your name will not be forgotten. Every death this technology can prevent will be laid at your door.” Correspondence obtained by Reuters through public-records requests shows that officials in Sweden and Norway have received dozens of emails and messages from Tesla enthusiasts citing claims about FSD’s life-saving capabilities. Other European governments denied or haven’t responded to Reuters requests seeking communications on FSD received from the public.

One August 2025 email to Swedish road regulators from a Tesla Model 3 owner said: “Not approving this means that more people die unnecessarily” and asked why regulators hadn’t approved FSD “when the statistics are so clear.” A group of Norwegian Tesla owners in May this year wrote an email to Norwegian regulators criticizing what they called “the passive and bureaucratic way” officials there handled deliberations over FSD, “instead of showing the necessary leadership and acting quickly to save lives.”

Stein-Helge Mundal, a chief engineer at the Norwegian Public Roads Administration, wrote in a November 2025 email to counterparts in Sweden, Finland and Denmark that Tesla was “advising their customers to contact RDW and their national authorities,” after a Tesla account on X urged customers to contact European regulators and linked to a contact form for Dutch regulator RDW. Mundal wrote to fellow regulators that, in light of Tesla’s “pressure” strategy, “we will need to use a lot of effort to answer misled consumers.”

In a statement to Reuters, Mundal said: “We do not approve systems because they are popular, and we do not reject them because they are controversial.” He added that ‌the term “Full Self-Driving” could lead drivers to assume the system “can do more than it actually can.” Norwegian regulators consider statements about lives saved to be “marketing claims,” he said.

Hans Nordin, an investigator ‌at the Swedish Transport Agency, declined to comment on Tesla’s pressure campaign or the safety data it presented. Denmark has approved FSD. Sweden and Finland haven’t announced a decision.

In France, minister Tabarot signaled a change of heart last month. He posted a photo of himself in a virtual meeting with a smiling Musk, describing a “constructive exchange” about FSD. He described “working closely” with Tesla on changes “that will enable France to support its approval.” In a statement to Reuters, the French transportation ministry didn’t respond to the criticism from Musk and other X users, or ​explain what led to Tabarot’s more optimistic tone.