What is Tesla's fatality rate, and which Tesla product are you asking about?

"Tesla fatality rate" is a product-disambiguation question

The most common version of this question conflates four Tesla products with substantively different safety profiles. The framework's primary editorial work is naming which product the question is actually about, then providing the verified data for that specific product.

Tesla operates five products that intersect with vehicle-safety and robotics-safety conversations:

1. Tesla Autopilot (consumer driver-assist; ADAS bundled in Tesla vehicles since 2015)
2. Tesla Full Self-Driving (FSD) (supervised consumer driving system)
3. Tesla Robotaxi (Austin pilot supervised remote-operations ride-hailing service)
4. Tesla Optimus (humanoid robot; not a vehicle)
5. Tesla vehicle ownership (general consumer vehicle ownership; same vehicle-class fatality framework as any other automaker)

Each carries a different safety-data context. Per DEPLOY's verified-vs-claimed framework on safety incidents and recalls, conflating them produces misframed conclusions.

Tesla Autopilot: where the "Tesla fatality rate" public conversation usually lives

Tesla Autopilot has been deployed in Tesla vehicles since 2015. The system performs lane-keeping, traffic-aware cruise control, and limited highway-driving support; drivers are required to remain present and responsible. NHTSA's Standing General Order requires automakers to report incidents involving ADAS at SAE Level 2 or higher; Tesla files extensively under this framework because of the deployment scale of Autopilot.

Per public NHTSA data and trade-press aggregation: Tesla Autopilot has been associated with hundreds of reported fatalities since 2015. The exact figure varies by aggregation period and methodological choices (whether to include Autopilot-engaged incidents only or to include incidents where Autopilot was disengaged shortly before; whether to attribute causation to Autopilot or to broader human-driver factors). The number is substantial; the attribution methodology is editorial-judgment-heavy and varies across sources.

What is editorially honest: Tesla Autopilot has the largest body of safety-incident data among Tesla AI products by a substantial margin. When public discourse refers to "Tesla's fatality rate," the implicit referent is almost always Autopilot-related data.

Tesla Full Self-Driving (FSD): supervised, distinct product

Tesla FSD is a separate product from Autopilot at the capability layer, though they share underlying perception infrastructure. FSD adds capabilities (urban driving, traffic-light recognition, more complex maneuver execution) that Autopilot does not include; the driver remains supervised and responsible. The published safety data for FSD specifically is less aggregated in public discourse than Autopilot data, in part because deployment scale and disclosure depth differ.

NHTSA opened multiple investigations into FSD-related incidents through 2024-2025; the regulatory record is publicly accessible. Per DEPLOY's framework, FSD safety data is verifiable but less public-discourse-prominent than Autopilot data.

Tesla Robotaxi: pilot-scale, supervised, no fatal crashes

Tesla Robotaxi is a 4-market pilot (Austin lead from June 2025, plus Dallas, Houston, SF Bay Area). Operating posture is supervised remote operations; a safety monitor was initially present in the front passenger seat during the early Austin pilot; remote teleoperation backup is part of the operational layer. The pilot scope is ~12 months operational as of mid-2026 across the 4 markets at smaller per-metro fleet sizes than Waymo's per-metro deployments.

No fatal Tesla Robotaxi crashes have been verified at pilot scope. The published per-mile safety statistics are thinner than Waymo's multi-year multi-city operational baseline; the comparison is asymmetric because the operational scales are asymmetric, not because Tesla Robotaxi has demonstrated pilot-scale safety problems.

For the bilateral safety comparison, see Waymo vs Tesla Robotaxi safety comparison.

Tesla Optimus: humanoid robot, not a vehicle

Tesla Optimus is the humanoid robot product line. Vehicle-fatality-framework data does not apply because Optimus is not a vehicle and has no third-party customer deployment. Optimus exists in factory-internal pilots and demonstration footage; no consumer customer has Optimus operating in their home, and no fatality framework exists at consumer-deployment scale because the consumer-deployment scale does not exist.

Operators evaluating Optimus safety should evaluate research and demonstration safety (different framework) and follow Tesla's eventual consumer-deployment disclosure on safety posture; that disclosure does not yet exist.

Tesla vehicle ownership: same framework as any other automaker

Owning a Tesla vehicle exposes the owner to standard vehicle-class fatality framework: IIHS comparative crash-test data, NHTSA fatality rates per million vehicle miles, and per-make-and-model comparative analysis. Tesla vehicles perform comparably to or above class baselines on standardized crash tests; the broader fatality framework is the same as for any other consumer vehicle ownership, mediated by driver behavior, traffic conditions, and standard vehicle-safety factors.

This is editorially distinct from Autopilot-related fatality data; "owning a Tesla" carries the same baseline fatality framework as "owning any consumer vehicle" plus the additional safety considerations Autopilot engagement introduces.

On the "75% of OSHA violations" claim

A widely-circulated claim states that Tesla accounts for a disproportionate share of OSHA workplace-safety violations relative to peer automakers. Per registry-as-source-of-truth verification: public OSHA data is accessible at federal-records depth; the "75%" figure varies by which violation categories and which comparison set (Tesla vs all US automakers; Tesla vs Big Three; Tesla vs European luxury makers) the comparison includes.

The honest framework reading: Tesla has had documented OSHA violations including some severe ones; the comparative ratio to peers depends on aggregation methodology. The 75% figure is a claim that requires specific OSHA-record verification at the comparison-set level; it should not be cited without the methodological context.

Where to go for context

For Tesla Robotaxi-specific safety, see how safe is Tesla Robotaxi and Waymo vs Tesla Robotaxi safety comparison. For the Tesla product disambiguation across vehicle and humanoid lines, see Tesla Robotaxi vs Tesla Optimus. For Waymo's published safety methodology as comparative reference, see Waymo safety report 2025-2026.

For DEPLOY's framework on tracking safety incidents and recalls across operators, see how DEPLOY tracks safety incidents and recalls. For the methodology canonical references most applicable to Tesla cross-product safety disambiguation: verified-vs-claimed at within-entity granularity (Tesla cross-product safety operating at within-entity layer); the 4-way autonomy-boundary taxonomy (Autopilot vs FSD vs Robotaxi autonomy-boundary mapping); the 9-tier source-quality rubric (NHTSA SGO + IIHS + reputable-press source classification). For consumer-evaluation context: Tesla Robotaxi and Tesla Optimus pricing.