FSD Recall Expansion Fuels Short Seller Claim That Tesla Robotaxi Story Is Over | Taha Abbasi

A Short Seller’s Bold Claim That FSD Recalls Will Kill Tesla’s Robotaxi Ambitions

GLJ Research, the investment firm run by noted Tesla bear Gordon Johnson, is making waves again with an aggressive short position on TSLA and a claim that should not be dismissed without examination. Johnson argues that the expanding pattern of FSD-related recalls by NHTSA effectively ends Tesla’s robotaxi story. His thesis is straightforward: a company that keeps getting forced to recall its driver-assistance software cannot credibly claim it is close to deploying fully autonomous vehicles without any human supervision.

It is easy to write off short sellers as motivated by their financial positions. That is often fair. But the underlying regulatory data Johnson is pointing to is real, and it raises legitimate questions about the gap between Tesla’s robotaxi promises and the current state of FSD performance.

The Recall History Is Getting Harder to Ignore

Tesla has issued multiple recalls related to its Autopilot and FSD systems over the past three years. The most significant came in December 2023, when Tesla recalled over 2 million vehicles after a two-year NHTSA investigation found that Autopilot’s driver monitoring was insufficient to prevent misuse. That recall, delivered as an over-the-air software update, added more aggressive driver attention reminders and limited the conditions under which Autopilot could be engaged.

Since then, additional recalls have addressed specific FSD behaviors, including issues with the system’s handling of certain intersection types, speed limit recognition, and responses to emergency vehicles. Each recall individually might seem minor. Software gets updated, the issue gets patched, and vehicles continue operating. But the cumulative pattern tells a story about a system that is still encountering failure modes that regulators consider serious enough to trigger formal action.

Johnson’s argument is that this pattern is fundamentally incompatible with the level of reliability required for unsupervised autonomous driving. A robotaxi service requires a system that can handle every driving scenario it encounters without human intervention. A system that is still subject to recalls for basic driving behaviors is, by definition, not at that level yet.

The Regulatory Framework Is Tightening

NHTSA’s approach to autonomous vehicle oversight has been evolving rapidly. The agency has moved from a largely hands-off posture during the early years of Autopilot to an increasingly active role in monitoring and regulating driver-assistance systems. The Standing General Order issued in 2021, which requires manufacturers to report crashes involving automated driving systems, has generated a significant database of incident data that the agency uses to identify potential defects.

For Tesla, the challenge is that FSD occupies an unusual regulatory space. It is the most widely deployed advanced driver-assistance system in the world, with millions of vehicles running the software in real-world conditions. That scale generates enormous amounts of data, which is a massive advantage for training and improving the system. But it also means that even rare failure modes affect large numbers of vehicles, which makes them visible to regulators in ways that smaller deployments can avoid.

Waymo, by contrast, operates a much smaller fleet of purpose-built vehicles in carefully mapped and geofenced areas. That approach reduces the range of scenarios the system encounters, which in turn reduces the likelihood of the kind of broad-spectrum failures that trigger recalls. Tesla’s approach of deploying FSD on consumer vehicles across the entire country is bolder and potentially more transformative, but it also creates more regulatory exposure.

Does This Actually End the Robotaxi Story?

Johnson’s conclusion that the recall pattern “ends the robotaxi story” is provocative, but it overstates the case. Here is why.

First, recalls are a standard part of the automotive industry. Every major manufacturer issues them regularly. The existence of recalls does not mean a product is fundamentally flawed. It means the manufacturer and the regulator are identifying and addressing issues, which is exactly how the safety process is supposed to work.

Second, the recalls Tesla has faced are software recalls, not hardware recalls. They are delivered over the air, often within days of being announced. This is fundamentally different from a mechanical recall that requires a physical visit to a dealership. The speed and efficiency of Tesla’s recall response process is actually one of the company’s strengths, even if the frequency of recalls is a legitimate concern.

Third, the robotaxi program does not depend on the current consumer version of FSD being perfect today. Tesla is developing the Cybercab as a purpose-built autonomous vehicle that will run a more advanced version of the FSD software, potentially with additional hardware capabilities. The consumer FSD product and the robotaxi product are related but not identical. Improvements from the consumer fleet feed into the robotaxi development, but the two products will not necessarily launch with the same software configuration.

The Market’s Reaction

Tesla’s stock price has been volatile in 2026, with the shares trading around $380 as of late March. The bull case for TSLA has long been built on the robotaxi opportunity, with analysts like Cathie Wood at ARK Invest valuing the company primarily on its potential to capture a massive share of the autonomous mobility market. If Johnson is right that the regulatory environment makes robotaxi deployment unlikely in the near term, a significant portion of Tesla’s current valuation is at risk.

But the market has heard the bear case before. GLJ Research has maintained a short position on TSLA for years, and the stock has repeatedly defied Johnson’s price targets. That history does not make his current analysis wrong, but it does provide context for evaluating the credibility and track record of the source.

The more nuanced reality is that Tesla’s robotaxi timeline is almost certainly going to slip beyond what Musk has publicly suggested. The June 2026 launch target for Austin is ambitious, and the regulatory approvals required to operate vehicles without human safety drivers are not guaranteed on any specific timeline. But “delayed” and “dead” are very different things.

What Investors Should Watch

For investors trying to evaluate the robotaxi thesis, the key metrics to track are not the recall count or Gordon Johnson’s price target. They are the specific technical and regulatory milestones that Tesla needs to hit before commercial deployment becomes possible.

Watch for NHTSA’s response to Tesla’s application for autonomous vehicle exemptions. Watch for the results of Cybercab testing in Austin and other markets. Watch for the FSD safety data that Tesla publishes quarterly, particularly the miles-per-intervention metric that serves as the best proxy for system reliability. And watch for any formal regulatory framework at the state or federal level that explicitly addresses the deployment of vehicles without human controls.

The FSD recall expansion is a data point, not a verdict. It reflects the growing pains of a system that is being developed at unprecedented scale. Whether those growing pains are a sign of fundamental inadequacy or a normal part of the path to full autonomy is the question that will determine whether the bulls or the bears are ultimately right.

Taha Abbasi covers autonomous vehicles, electric mobility, and frontier technology. For real-world testing of FSD and driverless systems, follow him on YouTube.

Related reading: Tesla FSD Cybertruck Crash Debate