Last reviewed
May 20, 2026
Sources
No citations yet
Review status
Needs citations
Revision
v1 ยท 5,255 words
Add missing citations, update stale details, or suggest a clearer explanation.
Cruise (legally Cruise LLC, formerly Cruise Automation) was an American self-driving vehicle company that operated as a subsidiary of General Motors from 2016 until early 2025, when GM acquired full ownership and discontinued its robotaxi business. Founded in San Francisco in October 2013 by Kyle Vogt and Daniel Kan, Cruise pivoted from aftermarket driver-assistance kits to fully driverless ride-hail and at its peak operated a paid, no-safety-driver robotaxi service in multiple U.S. cities.[1][2] After an October 2023 collision in which one of its autonomous vehicles dragged a pedestrian roughly 20 feet, the California Department of Motor Vehicles suspended Cruise's driverless permit and the company halted nationwide operations.[3][4] On December 10, 2024, GM announced it would stop funding the robotaxi business and would instead reallocate Cruise technology into advanced driver-assistance systems for personal vehicles; the subsidiary was folded into GM, roughly half its staff were laid off in February 2025, and the standalone robotaxi service never resumed.[5][6][7]
| Item | Detail |
|---|---|
| Legal name | Cruise LLC (formerly Cruise Automation, Inc.) |
| Founded | October 2013, San Francisco, California[1] |
| Founders | Kyle Vogt, Daniel Kan[1][2] |
| Parent | General Motors (subsidiary March 2016; wholly owned February 2025)[8][7] |
| Peak valuation | approximately $30 billion (January 2021)[9] |
| Cumulative GM operating losses | over $10 billion (2016 through 2024)[5] |
| Driverless permit (California) | issued June 2, 2022; suspended October 24, 2023[10][4] |
| Cities served (driverless) | San Francisco, Phoenix, Austin, Houston[11][12] |
| Last CEOs | Kyle Vogt (until November 2023); Marc Whitten (July 2024 to February 2025)[13][14][6] |
| Status (May 2026) | Robotaxi service permanently shut down; technology absorbed into GM's ADAS engineering[7][6] |
Kyle Vogt, a co-founder of the streaming companies Justin.tv and Twitch and an MIT alumnus who had worked on the DARPA Urban Challenge as a college student, incorporated Cruise Automation in October 2013 with Daniel Kan, another technology entrepreneur who had previously co-founded the on-demand laundry service Exec.[1][2] The startup was based in San Francisco and joined Y Combinator's Winter 2014 batch.[2] Its first product was the RP-1, an aftermarket retrofit kit intended to add highway-only autonomous capabilities to existing cars, demonstrated first on an Audi A4 and S4.[1] The kit consisted of a roof-mounted sensor package and an actuator module that could control the steering wheel, throttle and brake. The team quickly recognized that supporting the long tail of vehicle makes and models would be commercially infeasible, and by 2014 had pivoted to testing fully autonomous prototypes built around Nissan Leaf hatchbacks, then autonomous retrofits of the Chevrolet Bolt electric vehicle.[1] Cruise received its first California autonomous-vehicle testing permit in 2015.[1]
On March 11, 2016, General Motors announced the acquisition of Cruise Automation.[8] At the time of announcement, press reports described the deal as worth more than $1 billion in cash and stock; GM's subsequent securities filings showed closing consideration of $581 million, with $291 million in cash and approximately $290 million in newly issued GM common stock, plus contingent payments tied to milestones that lifted the headline figure.[15] Cruise was approximately a forty-person company at closing.[1] GM kept Cruise headquartered in San Francisco and operated it as an independent subsidiary with its own engineering culture, while folding it administratively into GM's autonomous vehicle and electrification group. Kyle Vogt initially served as president and chief technology officer; Dan Ammann, GM's president, took over as Cruise's chief executive officer in November 2018, with Vogt continuing in technology and product roles.[9]
Between 2018 and 2022 Cruise raised approximately $7 billion of external capital on top of GM's own commitments, becoming one of the most heavily funded private companies in the autonomous-vehicle sector. The principal rounds were:
The $30 billion January 2021 mark represented Cruise's peak valuation; it was never publicly remarked higher and was effectively retired when GM took the unit private in 2025.[7]
Cruise concentrated its on-road operations in San Francisco starting in 2017, choosing the city for its dense urban traffic, complex intersections and California's open testing regime.[1] Engineering build-out followed three steps. First, on-road testing with safety drivers using fleets of camera-, lidar- and radar-equipped Chevrolet Bolts produced petabytes of perception and planning data.[19] Second, in October 2020 Cruise received a California DMV permit to remove the human safety driver from its vehicles for limited testing without passengers, joining a small handful of operators authorized for fully driverless testing in California.[1] Third, in early 2022 Cruise began offering free, no-safety-driver rides to the public in limited zones of San Francisco using its Chevy Bolt fleet.[10]
On June 2, 2022, the California Public Utilities Commission issued the first Driverless Autonomous Vehicle Passenger Service Deployment permit in California to Cruise, permitting paid driverless rides for the public in a defined area of San Francisco between 10 p.m. and 6 a.m. at a maximum speed of 30 mph, using up to thirty all-electric autonomous Bolts.[10] Cruise began collecting fares almost immediately, making it the first company in a major U.S. city to commercially offer rides to the general public without a human driver.[10] In August 2023, California regulators expanded both Cruise's and Waymo's operating windows in San Francisco, although by that point Cruise had begun to attract negative attention for stalled vehicles blocking intersections and obstructing emergency responders.[20]
In September 2022, Vogt, by then back as chief executive officer, announced that Cruise would launch driverless ride-hail in Phoenix and Austin within ninety days.[11] Phoenix operations followed an existing Walmart delivery partnership; Austin represented the first city where Cruise had no pre-existing mapping or fleet infrastructure. Both cities began offering driverless rides to the public before the end of 2022.[11] During 2023, Cruise opened supervised testing in Houston, Dallas, Miami and Nashville, with an internal target of operating in ten cities by year-end.[12] On October 12, 2023, Cruise opened public driverless rides in Houston, operating from 9 p.m. to 6 a.m. in an 11-square-mile area covering downtown, midtown, East Downtown, Montrose, Hyde Park and River Oaks neighborhoods, with promotional $5 flat fares.[21] Cruise reported that its combined fleet across Austin, Houston, Phoenix and San Francisco numbered roughly 400 vehicles at the time, and that Vogt had told the company it had delivered more than 250,000 driverless rides cumulatively across all cities by late 2023.[21][13]
At approximately 9:30 p.m. on October 2, 2023, near the intersection of Fifth and Market Streets in San Francisco, a hit-and-run driver in an adjacent lane struck a pedestrian, propelling her into the path of a Cruise Chevrolet Bolt operating without a driver, internally nicknamed "Panini."[3][22] The Cruise vehicle braked but could not avoid contact and stopped over the pedestrian. After the initial stop, the Cruise vehicle resumed motion at approximately 7 mph in a programmed pullover maneuver and dragged the pedestrian roughly 20 feet before coming to a final halt with her still trapped beneath one wheel.[3][22] First responders required hydraulic rescue tools to extract her, and she was transported to San Francisco General Hospital with severe injuries.[3]
The DMV asserted that when Cruise representatives presented the agency with video on October 3, they showed footage that ended with the vehicle's initial stop and did not include the subsequent pullover and dragging.[4] Cruise disputed this characterization, saying the full video had been shown during the same meeting.[4] An independent investigation commissioned by Cruise's board and conducted by the law firm Quinn Emanuel Urquhart & Sullivan reviewed more than 200,000 documents and interviewed 88 Cruise employees; the resulting 195-page report concluded that "deficient leadership" at the highest levels of the company, including senior executives and the legal department, "poor leadership, mistakes in judgment, lack of coordination," and an "us versus them mentality with government officials" had driven Cruise's flawed response.[23] Among the report's findings: a paralegal had been delegated responsibility for filing the initial federal report on the incident.[23]
On October 24, 2023, the California DMV suspended Cruise's driverless deployment and driverless testing permits effective immediately, citing public-safety concerns and stating that the company had "misrepresented" the safety of its technology and that the vehicles were "not safe for the public's operation."[4] Cruise's permit for testing with a safety driver was unaffected.[4] Three days later, on October 27, 2023, Cruise voluntarily paused all driverless operations across every city in which it operated.[3] On November 8, 2023, Cruise issued a recall via software update for 950 of its autonomous Bolts addressing the post-collision behavior that had caused the dragging.[22]
CEO and co-founder Kyle Vogt resigned on the evening of November 19, 2023. He cited the company's accomplishments, including more than 250,000 cumulative driverless rides, but offered no successor.[13] Mo Elshenawy, executive vice president of engineering, became president and chief technology officer; Craig Glidden, an existing GM executive, took on chief administrative officer responsibilities; Jon McNeill joined as vice chairman of the Cruise board.[13] The following day, Vogt's co-founder Daniel Kan also stepped down.[13]
On December 13, 2023, Cruise dismissed nine senior leaders, including its chief operating officer and head of government affairs. The next day, December 14, 2023, the company laid off 900 employees, or roughly 24 percent of its workforce, with most cuts concentrated in commercial operations and corporate staffing rather than engineering.[24] Departing employees were placed on payroll through February 12, 2024, with eight weeks of additional pay, two extra weeks per year of tenure up to three years, and health benefits through May. Cruise's headcount fell from roughly 3,800 to around 2,900.[24]
The custom-built Cruise Origin shuttle, which had been preparing for production at GM's Factory Zero in Detroit, was paused in November 2023 and then indefinitely shelved in July 2024 in favor of a return to the Chevrolet Bolt platform.[25] GM took a write-down on Origin-related inventory as part of its end-of-year accounting.[26]
Cruise began a measured restart in 2024. In April it sent manually driven Bolts onto public roads in Phoenix to refresh its high-definition maps, and in May 2024 it began supervised autonomous testing with a human safety operator behind the wheel in Phoenix, then Dallas and Houston.[27][28][29] On June 25, 2024, GM and Cruise named Marc Whitten, a longtime technology executive who had been a founding engineer on Microsoft's Xbox and Xbox Live, a senior vice president at Amazon for Fire TV, Kindle and Echo devices, chief product officer at Sonos and chief product and technology officer for Unity Create, as Cruise's new chief executive officer.[14][30] Whitten started on July 16, 2024, charged with rebuilding safety culture and the path back to driverless commercial operations.[14]
In September 2024 (announced October 1, 2024), the National Highway Traffic Safety Administration imposed a consent order with a $1.5 million monetary penalty on Cruise for failing to disclose, in two reports under NHTSA's Standing General Order for automated-driving-system crashes, the post-collision dragging in the October 2 incident.[31] The order required Cruise to submit a corrective action plan and meet quarterly with NHTSA. A separate Department of Justice agreement saw Cruise admit to submitting a false report and pay a $500,000 fine to settle that federal investigation.[32]
On December 10, 2024, General Motors announced that it would no longer fund Cruise's robotaxi development. GM said it would "realign its autonomous driving strategy to prioritize the development of advanced driver assistance systems on a path to fully autonomous personal vehicles" and would combine Cruise's technical teams with its own engineering organization.[5] In an accompanying disclosure GM stated it expected annual run-rate savings of more than $1 billion when the restructuring was complete in the first half of 2025; GM had spent more than $10 billion on Cruise since the 2016 acquisition while generating less than $500 million of cumulative revenue from the robotaxi service.[5][7] On December 11, 2024, Honda separately said it would withdraw from the planned 2026 Tokyo robotaxi joint venture with GM and Cruise.[33]
On February 4, 2025, GM announced that it had completed the previously disclosed merger to acquire the remaining minority shares of Cruise from outside investors, taking ownership to 100 percent.[7] In the same week Cruise informed staff that approximately 50 percent of its workforce, around 1,000 positions, would be laid off; the cuts fell most heavily on non-technical functions, with GM stating that more than 80 percent of engineering and other technical roles would be retained and folded into GM's ADAS organization.[6] Marc Whitten and several other senior leaders departed during the restructuring.[6] As of May 2026, Cruise no longer operates a robotaxi service, the standalone Cruise brand has been effectively retired in favor of GM's Super Cruise driver-assistance product line, and the engineering team continues development of advanced driver-assistance systems and personal-vehicle autonomy under GM rather than as a separate subsidiary.[7][6]
Cruise pursued a vertically integrated, full-stack architecture in which the same organization designed the perception, prediction, planning and control software and integrated it with vehicle hardware in close cooperation with GM. The principal building blocks were:
The bulk of Cruise's commercial fleet ran on retrofitted Chevrolet Bolt electric vehicles fitted with a roof-mounted sensor pod, redundant compute hardware, drive-by-wire actuation and a tertiary safety system layered on top of the production Bolt's controls.[19] These vehicles formed the basis of the company's San Francisco, Phoenix, Austin and Houston commercial services. The purpose-built Cruise Origin was revealed at a January 21, 2020 event in San Francisco and was designed jointly with Honda and General Motors.[16][34] The Origin had no steering wheel, no pedals and no driver's seat; instead, two benches faced one another in a 3+3 layout with passenger sliding doors. It was electric and intended for purpose-built ride-hail and delivery service. GM produced a small number of pre-commercial Origins at Factory Zero in Detroit before pausing production in November 2023 and indefinitely shelving the vehicle in July 2024.[25]
Cruise's perception stack combined three complementary modalities: lidar, cameras and radar.[19] Cameras provided high-resolution color and texture information used for traffic-light state recognition, lane-marker detection and vehicle-light inference (for example, distinguishing hazard lights and turn signals to predict whether a parked vehicle was likely to remain stationary).[19] Lidar provided 3D structure for obstacle detection and reliable ranging in low-light conditions. Radar provided weather-robust velocity estimation. The system used late sensor fusion, in which obstacles were detected independently in each modality and then combined using extrinsic calibration parameters to project 2D camera detections into the 3D scene.[19]
Cruise's perception pipeline followed a recognizable autonomous-vehicle pattern: sensor calibration (intrinsic and extrinsic), time synchronization, active-learning-based labeling, fleet-data management and per-task models for detection and tracking.[19] For behavior prediction in busy urban scenes, the company used recurrent neural networks over sequences of perceived states to infer, for example, whether a vehicle was double-parked or merely stopped, and probabilistic trajectory prediction at intersections.[19]
The planning stack produced a desired trajectory from the predicted scene, and a model-predictive controller computed the steering angle and longitudinal acceleration commands needed to follow the trajectory while respecting actuator limits and comfort constraints.[19] Cruise relied on high-definition maps for localization, semantic scene priors and geofencing to constrain the operational design domain to areas where the company had validated its software.[19]
Cruise's January 2021 partnership with Microsoft designated Microsoft Azure as its preferred cloud platform for compute-intensive workloads including simulation, training and fleet data processing.[9] Internally, the company developed a visualization tool nicknamed Webviz, built on the Robot Operating System (ROS), to inspect raw and processed sensor outputs (lidar point clouds, radar tracks, camera frames) in synchronized timeline view; Webviz was open-sourced and reused widely in the broader robotics community.[19] Cruise's data infrastructure used Apache Spark for daily aggregation of terabytes of fleet logs supporting distributed training and querying.[19]
| Year | Event |
|---|---|
| 2013 | Cruise Automation founded by Kyle Vogt and Daniel Kan in San Francisco[1] |
| 2014 | Y Combinator W14 batch; pivot from RP-1 retrofit kit to integrated AV[2] |
| 2015 | First California DMV autonomous-vehicle testing permit[1] |
| March 2016 | Acquired by General Motors for ~$581M closing consideration plus contingent payments[8][15] |
| 2017 to 2020 | San Francisco testing with Chevrolet Bolt; California driverless-testing permit (October 2020)[1] |
| 2018 | SoftBank Vision Fund commits $2.25B; Honda commits $2.75B; Dan Ammann becomes CEO[9] |
| January 2020 | Cruise Origin shuttle revealed in San Francisco[16][34] |
| January 2021 | $2B round including Microsoft at ~$30B valuation; Azure partnership[9] |
| February 2022 | Kyle Vogt becomes CEO; SoftBank exits in March, GM share rises to ~80%[18] |
| June 2022 | First California CPUC permit for paid driverless ride-hail in San Francisco[10] |
| September 2022 | Driverless commercial launch in Phoenix and Austin within 90 days announced[11] |
| May 2023 | Supervised testing announced in Houston and Dallas[12] |
| August 2023 | SFFD complaints; California DMV orders SF fleet halved after incidents[20] |
| October 2, 2023 | Pedestrian dragged ~20 ft after being thrown into Cruise vehicle's path[3] |
| October 12, 2023 | Public robotaxi service opens in Houston[21] |
| October 24, 2023 | California DMV suspends driverless permits[4] |
| November 8, 2023 | Recall of 950 vehicles via software update[22] |
| November 19, 2023 | Kyle Vogt resigns; Dan Kan follows[13] |
| December 14, 2023 | 900 employees (~24%) laid off; nine senior leaders dismissed[24] |
| January 25, 2024 | Quinn Emanuel report blames "deficient leadership"[23] |
| May to June 2024 | Supervised testing resumes in Phoenix, Dallas and Houston[27][28] |
| July 16, 2024 | Marc Whitten begins as CEO[14] |
| October 1, 2024 | NHTSA $1.5M consent order; DOJ false-statement settlement[31][32] |
| December 10, 2024 | GM announces end of robotaxi funding; ADAS pivot[5] |
| February 4, 2025 | GM completes full acquisition; ~50% workforce laid off; Whitten departs[7][6] |
Cruise's contemporaries pursued a range of strategies that, in retrospect, illuminate why the company's trajectory diverged from those of survivors. A simplified comparison as of May 2026:
| Program | Approach | Status (May 2026) | Comment |
|---|---|---|---|
| Waymo | Full-stack L4 robotaxi; bespoke lidar; cautious city-by-city scale-up | Active commercial service in Phoenix, San Francisco, Los Angeles and additional U.S. metros; the only U.S. operator to scale paid driverless service through the period[35] | Scaled more slowly than Cruise pre-2023 but avoided permit-suspending incidents |
| Cruise | Full-stack L4 robotaxi; Chevrolet Bolt + Origin; rapid multi-city scale-up | Robotaxi service permanently shut down; technology absorbed into GM ADAS[6] | Aggressive expansion preceded the October 2023 collision and regulatory rupture |
| Zoox | Full-stack L4; purpose-built bidirectional vehicle; Amazon subsidiary | Limited commercial deployment in Las Vegas; expanding to additional cities | Pursues a similar shared-shuttle product to the shelved Cruise Origin |
| Wayve | "AV2.0" end-to-end learned driving foundation models (e.g. GAIA-2, LINGO-2); software-only | Pre-commercial; partnerships with global OEMs and Uber for future deployment | Argues that hand-engineered modular stacks like Cruise's do not scale; competes on data-driven generalization |
| Tesla (Robotaxi / FSD) | Vision-only end-to-end neural network on production vehicles; software updates over the air | Robotaxi launched in Austin in 2025 on Model Y / Cybercab; ongoing regulatory scrutiny of supervised FSD | Pursues a fundamentally different sensor philosophy (cameras only) from Cruise (lidar+camera+radar)[36] |
| Baidu Apollo Go | Full-stack L4 robotaxi in China; Apollo platform | Active commercial driverless service in multiple Chinese cities including Wuhan and Beijing | China's analogue to Waymo; demonstrates that robotaxi can scale under different regulatory regimes |
| Pony.ai | Full-stack L4; robotaxi and truck programs; U.S./China presence | Listed on Nasdaq November 2024; operating driverless ride-hail in Beijing, Guangzhou, Shenzhen and Shanghai | Smaller in scale than Waymo but operating without the kind of public-safety pause Cruise experienced |
| Mobileye | Tier-1 vision software supplier and L4 robotaxi platform (Mobileye Drive); spun out of Intel | Powers driver-assistance in tens of millions of vehicles; L4 pilots ongoing | Pursues primarily a supplier model rather than a Cruise-style operator model |
Of the U.S. L4 robotaxi operators that scaled to public, paid, no-safety-driver service before 2024, only Waymo continued to operate at the close of 2025. Cruise was the highest-profile shutdown but not the only one: Ford and Volkswagen's jointly funded Argo AI had already wound down in October 2022, citing similar capital-intensity concerns.[37]
Cruise's rise and fall has become a frequently cited case study in the autonomous-vehicle industry. Several themes emerge from public reporting and the Quinn Emanuel post-mortem:
Safety culture and disclosure. The Quinn Emanuel report attributed the company's regulatory collapse less to the technical failure of the October 2 pullover behavior than to the response after the collision, particularly the failure to clearly disclose the post-collision movement to the DMV and NHTSA. Both the California DMV's suspension order and NHTSA's consent order emphasized incomplete or inaccurate disclosure rather than the underlying technical defect alone.[4][31][23] The pattern is widely cited as a warning about the relationship between AV operators and regulators.
Capital intensity and timing. GM's December 2024 announcement framed the decision as one about capital allocation. The company had spent more than $10 billion through Cruise and projected that several billion more would be required to scale to profitability, and concluded that the cost and timeline did not justify continued standalone investment given the competitive position of Waymo and emerging Chinese operators.[5] Industry analysts cited Cruise's experience alongside Argo AI's earlier closure as evidence that robotaxi development was a more capital-intensive bet than incumbents had assumed.
Operational design domain. Public reporting in 2023 documented repeated incidents in which Cruise vehicles stalled, blocked intersections, obstructed emergency responders or behaved unexpectedly in atypical conditions such as fog or construction zones.[20] The pattern suggested that the company had expanded into more cities than its operational design domain comfortably supported, contributing both to public-safety risk and to political and regulatory friction.
Hardware-platform decisions. The decision to invest in the purpose-built Cruise Origin, with its sliding doors and steering-wheel-free interior, depended on the program scaling to volumes that justified bespoke manufacturing. When operations halted, the inventory and capital invested in Origin tooling at Factory Zero became a stranded cost, and the vehicle was indefinitely shelved.[25] By contrast, Waymo partnered with multiple existing OEM platforms (Chrysler Pacifica, Jaguar I-Pace, Zeekr) without committing to its own purpose-built vehicle until commercial service was established.
Integration with parent ADAS programs. GM's stated rationale for the 2025 absorption was that the engineering investment in perception, planning and verification was reusable for personal-vehicle advanced driver-assistance systems like Super Cruise.[7] As of May 2026, Super Cruise remained one of GM's principal product differentiators across more than twenty vehicle models, and former Cruise engineers were folded into the team developing eyes-off, hands-off variants.[7] The episode raises the question, debated by industry observers, of whether a Tier-1 ADAS path was the right destination for the technology all along or whether an opportunity to build a Waymo-scale robotaxi business was foreclosed by execution failures.
Several specific criticisms recurred across reporting on Cruise in the period leading up to and following October 2023:
The Cruise story sits in a wider context that includes: