US auto safety regulators opened a new inquiry into Tesla’s Full Self-Driving after documenting dozens of instances where cars allegedly blew red lights, went the wrong way, and violated traffic laws with the driver-assistance system engaged. The National Highway Traffic Safety Administration said it is aware of 58 such incidents and launched a preliminary evaluation covering about 2.9 million vehicles. Shares slipped as the latest probe adds to a growing list of regulatory fronts surrounding Autopilot, crash reporting, and Tesla’s broader autonomy push. The company did not immediately respond to a request for comment.
Scope of the new case – NHTSA’s filing outlines a discrete set of events tied to FSD behavior at intersections and on public roads, including six reports where Teslas entered an intersection against a red light and then crashed with other vehicles. Multiple incidents occurred at the same Maryland intersection; the agency said Tesla has taken steps to address that location. Importantly, NHTSA noted that none of the incidents now under review in this evaluation involve a fatality. The agency is assessing the scope, frequency, and safety consequences of the behaviors, with a focus on whether FSD detects and responds to traffic-control signals and roadway rules as designed when drivers are in supervisory mode. Tesla markets FSD as a partial-automation system that requires constant driver attention and hands-on readiness.
Market reaction – Tesla stock traded lower on the headlines, reflecting investors’ sensitivity to any threat to the company’s autonomy narrative and software margins. As of Thursday afternoon, shares were down about 1.4% to around 432.62, with an intraday range that underscored the headline risk around safety and regulatory actions. The latest dip follows a volatile stretch tied to competing storylines: enthusiasm around robotaxi ambitions on one hand, and intensifying scrutiny of driver-assistance performance on the other. For a name that carries a premium for perceived leadership in AI-enabled driving, more safety noise typically compresses the multiple, at least until the regulatory picture clears or Tesla demonstrates durable software improvements.
A pattern of scrutiny – Today’s case is not happening in isolation. NHTSA’s Office of Defects Investigation is already probing whether FSD can detect and appropriately respond to reduced visibility conditions such as fog, sun glare, or airborne dust, a separate inquiry opened after a fatal crash. Autopilot, the company’s highway-focused driver-assistance system, is under a broad review tied to 956 crashes reported between 2018 and 2023, including 29 fatalities, as regulators assess driver monitoring and misuse. Beyond software, NHTSA has active workstreams on issues ranging from door latches to Tesla’s timeliness in reporting collisions. The through line: Washington is stress-testing whether Tesla’s automation stack performs safely and predictably, and whether oversight mechanisms keep pace with fast-deployed over-the-air updates.
What NHTSA wants to know – Regulators are zeroing in on two practical questions that matter to safety and liability. First, does FSD reliably obey traffic control devices, lane markings, and right-of-way rules in complex environments like unprotected turns and multi-lane intersections? Second, does the system remain robust under degraded conditions—low visibility, glare, or partial occlusion—that can confound sensors and software? The wrong-way and red-light allegations go right to the heart of that inquiry. There is also an expectations gap. Tesla says drivers must supervise at all times; Elon Musk has suggested that in some states drivers may soon not need to pay attention when FSD is engaged. That tension sets the stage for regulators to demand clearer guardrails or more conservative feature behavior.
Recall risk and the ODI ladder – A preliminary evaluation can escalate to an engineering analysis and ultimately a recall if defects are found. For software-led vehicles, a recall often means mandatory over-the-air changes, as Tesla did in 2023 by bolstering driver monitoring and engagement prompts around Autosteer. This time, potential remedies could include stricter speed and intersection controls, tighter driver-attention enforcement through camera-based monitoring, geofencing in known-problem locales, or disabling certain maneuvers until the system meets a higher performance bar. Any material curbs to FSD functionality would weigh on the company’s promise of high-margin software revenue. Tesla has sold FSD packages for thousands of dollars per vehicle and recognizes the revenue over time as features are delivered. Slower delivery or feature rollbacks can push out recognition and dull attach rates.
Musk’s autonomy timeline vs. Washington – The new probe lands just as Tesla pitches its next act: a steering wheel-free Cybercab robotaxi Musk says could be in production by 2026. That vision presumes not only technical readiness but also regulatory comfort with vehicles lacking traditional controls under Federal Motor Vehicle Safety Standards. The more NHTSA documents basic traffic-law noncompliance in supervised FSD, the harder it is to argue for rapid approvals of fully driverless variants. Analysts have already flagged a mismatch between Tesla’s aggressive autonomy timelines and the incrementalist stance regulators take after high-profile incidents. Today’s case strengthens the argument that any robotaxi rollout will be staged, geo-fenced, and scrutinized—if it stays on the current schedule at all.
The valuation hinge – Tesla trades on more than cars; the stock reflects a software and AI option value that assumes rising take-up of autonomy features and a pathway to robotaxi economics. That optionality is sensitive to headline risk, because every probe threatens to add costs, delay feature releases, or impose compliance overhead that eats into software margins. If NHTSA’s investigation results in broader limitations on FSD usage or new monitoring mandates, investors will mark down the near-term software contribution and focus back on hardware mix, pricing, and cost efficiencies. On the other hand, if Tesla can show measurable safety gains, particularly in adverse conditions and at intersections, it can rebuild confidence that the autonomy story remains intact despite more guardrails.
What to watch next – The next milestones are procedural but market-moving. Expect NHTSA to issue formal information requests asking for incident logs, video, and software version histories tied to the 58 events—and potentially beyond. Tesla’s response cadence and the specificity of its proposed mitigations will signal whether this stays a contained fix or expands into a broader engineering analysis. Watch for software release notes that adjust intersection handling, red-light detection thresholds, or driver monitoring. States may also weigh in, especially where incidents cluster, adding local restrictions even ahead of federal action. For investors, the tell will be whether the agency’s language drifts from performance questions to defect findings. If that happens, odds of a recall-grade OTA update rise—and with them, short-term pressure on TSLA’s software premium.
