Tesla Disabled “Rolling Stop” Functionality Associated with the “Aggressive” Driving Mode

The “rolling stop” functionality within the “Aggressive” Full Self Driving (FSD) profile that was released via a Tesla firmware update was recalled and disabled.

Risk classification

• Primary risk domain: 7 AI system safety, failures, & limitations
• Primary risk subdomain: 7.3 Lack of capability or robustness

The FSD system's rolling stop feature represents a lack of capability or robustness to safely and legally navigate intersections, creating a safety risk that prompted a regulatory recall.

Additional risk subdomains

• 5.1 Overreliance and unsafe use: The FSD profiles allowed users to select 'Assertive' mode, which encouraged unsafe driving behaviors like smaller following distances and illegal rolling stops.

Causal factors

• Entity: AI
• Intent: Intentional
• Timing: Post-deployment

The risk is presented as the AI system's autonomous driving behavior, specifically its decision to perform rolling stops at intersections.

EU AI Act risk tier

• Risk tier: 2 High Risk

High Risk: AI systems used in critical infrastructure, such as road traffic management and autonomous vehicles, are classified as High Risk under the EU AI Act due to their significant implications for public safety.

AI system and alleged parties

• AI system: Full Self-Driving (FSD) Beta (Tesla)
• AI purpose: Autonomous Driving; Navigation Assistant
• Behaviour type: Autonomous
• Alleged developer: Tesla
• Alleged deployer: Tesla
• Alleged harmed parties: Tesla drivers

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Negligible. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

People affected

• Occurrences reported: 1
• People reportedly exposed: 53822

Potential causes

Management

• Prioritizing growth over safety: Management pushed out FSD beta updates despite unresolved safety issues.
• Programming illegal maneuvers: Decision to program vehicles to violate traffic laws for user comfort.
• Monetization of beta software: Increasing prices to 12,000 dollars while still in beta.

Technology

• Rolling stop software design: Software allowed vehicles to travel through stop signs up to 5.6 mph.
• Aggressive driving profiles: Profiles like Assertive reduced following distance and increased risk.
• Flawed FSD beta software updates: Beta 10.3 released with bugs in left turns and unexpected stopping.

Data Inputs

• Object detection reliance: System relied on detecting pedestrians, cyclists, and cars to roll stops.
• Visibility and speed inputs: Relied on road speed limit and visibility data to trigger rolling stops.
• Safety Score system data: System prioritized drivers based on controversial safety metrics.

Human Factors

• Driver selection of risky modes: Drivers could select Assertive mode, choosing convenience over safety.
• Misleading marketing names: Terms like 'Full Self-Driving' cause driver over-reliance on Level 2.
• Over-reliance on Level 2 system: Drivers treating driver-assistance features as fully autonomous.

Process and Methods

• Prioritizing rider preference: System designed to mimic human errors like rolling stops for convenience.
• Public road beta testing: Testing unproven software with a large consumer fleet on public roads.
• Lack of rigorous pre-release testing: Releasing updates like 10.3 that had to be immediately pulled.

Regulatory Environment

• Lack of proactive AV standards: No regulations prevented programming vehicles to perform illegal stops.
• Delayed regulatory action: NHTSA intervened only after 53,822 vehicles received the software.
• Lack of federal AV oversight: Regulators allowed beta expansion before resolving basic safety issues.

Information quality

• Classification confidence: High
• Reason for confidence: The reports include official NHTSA recall documentation which provides precise figures (53,822 vehicles) and clear technical conditions under which the rolling stop feature activated. Media reports corroborate the details of the FSD profiles ('Chill', 'Average', 'Assertive') and the regulatory response, leaving very little ambiguity about the nature of the incident.

Tesla recalled its Full Self-Driving Beta software to disable a rolling stop feature that allowed vehicles to cross intersections without stopping, violating traffic laws. The recall affected over 53,000 vehicles but resulted in no crashes or injuries, presenting minor physical security risks that were managed through standard regulatory processes.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: No clear primary
• Alleged perpetrator: Tesla

Threat characteristics

• Imminence: Long-term. The issue was voluntarily resolved via an over-the-air firmware update in February 2022, posing no active or immediate threat.
• Autonomy: Human-supervised. While the FSD system operates the vehicle autonomously, it requires active human driver supervision and intervention capability.
• Novelty: Established threat. Software recalls and updates for driver-assist and autonomous vehicle systems are established procedures in the automotive industry.

Impact by dimension

• Physical security: Minor. The FSD feature allowed vehicles to autonomously perform illegal rolling stops, presenting a potential physical safety risk. However, no crashes or injuries occurred, and the issue was resolved via standard regulatory recall and software update.
• Information security: Negligible. No information warfare, intelligence compromise, or systematic information manipulation occurred in this incident.
• Sovereignty: Negligible. No threat to state authority or government operations. Standard regulatory bodies successfully engaged with the manufacturer to resolve the issue.
• Economic security: Negligible. No significant threat to economic stability, strategic industries, or national technological competitive advantage.
• Societal stability: Negligible. No threats to societal stability, civil liberties, or population-scale safety were identified in this commercial software recall.