Tesla Robotaxi Crashes, GM AI Hiring Shift

Tesla disclosed details on 17 robotaxi crashes — including two caused by remote human operators — in newly unredacted filings with the National Highway Traffic Safety Administration, while General Motors cut more than 600 IT workers and began replacing them with AI-focused hires. The disclosures and workforce moves, reported this week by Wired and TechCrunch, highlight how deeply human decisions still shape autonomous vehicle outcomes — and how aggressively automakers are restructuring around AI capabilities.

Tesla’s Teleoperator Crashes Reveal a Hidden Safety Layer

Both teleoperator-caused incidents happened in Austin, Texas, and both involved safety monitors riding in the passenger seat with no paying customers aboard. According to TechCrunch, the first crash occurred in July 2025, when Tesla’s automated driving system stalled on a street and the safety monitor requested remote assistance. A teleoperator took control, turned the vehicle left, drove it up a curb, and struck a metal fence at 8 mph. The monitor reported minor injuries but was not hospitalized.

The second incident occurred in January 2026. Again, a safety monitor requested navigation help from Tesla’s remote team. The teleoperator drove the car directly into a temporary construction barricade at 9 mph, scraping the front left fender and tire. Tesla reported no injuries in that case.

As Wired noted, Tesla told lawmakers earlier this year that it permits remote operators to drive vehicles at speeds up to 10 mph — a capability the company described as a way to move stalled vehicles without waiting for field representatives. That explanation now sits alongside two confirmed crashes caused by that same capability.

Why Tesla’s NHTSA Filings Changed

Tesla had previously redacted all crash narrative descriptions in its NHTSA submissions, citing confidential business information. That changed this week: the NHTSA database now includes narrative descriptions for all 17 crashes Tesla recorded since launching its Austin robotaxi network in mid-2025, according to TechCrunch.

The reason for the policy reversal is not publicly known. Tesla, which disbanded its public relations team years ago, did not respond to Wired’s request for comment.

The broader dataset of 17 incidents spans July 2025 through March 2026. The two teleoperator crashes are notable because they shift liability attribution — the autonomous driving system was not at fault in either case. Instead, the crashes occurred during human-controlled interventions, a scenario that most public discussions of self-driving safety do not account for.

All U.S. self-driving operators maintain remote monitoring teams, according to letters submitted to a U.S. senator and cited by Wired. Tesla appears to be an outlier in how frequently it allows those remote workers to directly drive vehicles rather than simply advise safety monitors.

GM’s AI Hiring Swap: 600 Jobs Out, AI Roles In

General Motors laid off more than 10% of its IT department — approximately 600 salaried employees — in what the company described as a deliberate skills rebalancing, according to TechCrunch Mobility. GM said it is actively hiring to fill those positions with workers who have AI-focused backgrounds, though the exchange is not expected to be one-to-one.

The capabilities GM is recruiting for include:

• AI-native development — building systems from the ground up, not using AI as a bolt-on tool
• Data engineering and analytics
• Cloud-based engineering
• Agent and model development
• Prompt engineering
• New AI workflow design

The distinction GM is drawing — between workers who use AI tools and workers who build AI systems — reflects a broader shift in how large companies are structuring technical roles. The company wants engineers who design training pipelines and architect models, not those who use AI assistants to write code faster.

Automakers’ Broader Job Cuts

GM’s restructuring sits within a larger pattern across Detroit’s legacy automakers. TechCrunch Mobility cited CNBC calculations showing that Ford, GM, and Stellantis combined have cut more than 20,000 U.S. salaried jobs — roughly 19% of their combined workforces — from employment peaks earlier this decade.

The causes are varied, but technological change, including AI adoption, is a consistent thread across the cuts. The job losses are not uniform: some roles are being eliminated outright, while others are being replaced by different skill profiles.

One company cited as having found a concrete revenue-generating AI application is Samsara. According to TechCrunch Mobility, Samsara spent a decade collecting camera footage from inside millions of commercial trucks for driver monitoring and liability purposes. The company used that data to train a proprietary model capable of detecting potholes — a use case with direct commercial value for fleet operators managing vehicle maintenance costs.

The contrast between Samsara’s data-driven approach and the more diffuse AI strategies at larger automakers illustrates a recurring tension: companies with focused, high-volume data assets appear better positioned to extract measurable value from AI investments than those applying AI broadly across complex organizations.

Remote Operations Remain a Regulatory Blind Spot

The Tesla crash disclosures raise a question that current NHTSA reporting frameworks were not designed to answer cleanly: when a human teleoperator causes a crash in an otherwise autonomous vehicle, how should that incident be classified and who is responsible?

Existing NHTSA reporting requirements for autonomous vehicle incidents focus on the automated driving system’s behavior. Crashes caused during human remote-control interventions occupy an ambiguous category — they are neither traditional driver-at-fault incidents nor ADS failures. Tesla’s filings attribute both Austin crashes to teleoperator actions, which technically clears the ADS of fault, but the incidents still occurred within Tesla’s robotaxi operation.

As robotaxi networks scale beyond early pilot cities, the frequency of remote interventions is likely to increase alongside total miles driven. The two confirmed teleoperator crashes represent a small fraction of Tesla’s 17 reported incidents, but the underlying dynamic — humans stepping in when autonomous systems get stuck — is structural, not incidental.

What This Means

The week’s automotive AI news points in two directions simultaneously. Tesla’s NHTSA disclosures show that autonomous vehicle operations are more human-dependent than the term “self-driving” implies. Remote operators are not a temporary training-wheels phase; they are a permanent safety layer, and their actions carry real crash risk. The regulatory and liability frameworks for this layer are still catching up to operational reality.

On the workforce side, GM’s AI hiring swap is a concrete example of what industry-level AI adoption looks like in practice: not mass automation of products, but replacement of one type of technical worker with another. The net job loss is real — 600 out, fewer than 600 back in — and the skills gap between displaced IT workers and the AI-native roles being created is not easily bridged by retraining programs. Ford, GM, and Stellantis have collectively shed roughly one in five salaried U.S. jobs this decade. The pace of that reduction, and the AI-focused shape of the replacements, suggests the restructuring has further to go.

FAQ

What caused Tesla’s robotaxi crashes in Austin?

Two of Tesla’s 17 reported robotaxi crashes were caused by human teleoperators who took remote control of vehicles when the automated driving system encountered difficulties. In both cases, the remote driver struck a fixed object — a metal fence and a construction barricade — at speeds below 10 mph.

Why is GM laying off IT workers if it is also hiring for AI roles?

GM described the cuts as a deliberate skills swap, eliminating approximately 600 salaried IT positions to create budget and headcount for workers with AI-native backgrounds in areas like model development, data engineering, and AI workflow design. The company acknowledged the exchange will not be one-to-one, meaning a net reduction in IT headcount is expected.

Are all self-driving companies required to report crashes to NHTSA?

Yes, U.S. autonomous vehicle operators are required to submit crash reports to the National Highway Traffic Safety Administration. Tesla had previously redacted all narrative descriptions from its filings, citing confidential business information, but unredacted its full dataset of 17 incidents this week without publicly explaining the change.