The short answer: Not quite yet, but closer than ever before.

As we reach the halfway point of 2025, Tesla stands at a pivotal moment in its autonomous driving journey. After nearly a decade of ambitious promises, the company is finally preparing to launch its first truly “unsupervised” self-driving service—but with important caveats that reveal both progress and ongoing limitations.

The Current State: Still Supervised

Despite its name, Tesla’s “Full Self-Driving (Supervised)” remains exactly that—supervised. As of May 2025, all Tesla vehicles equipped with FSD still require an attentive driver with hands on the wheel, ready to intervene at any moment. The system represents Level 2 automation according to SAE standards, meaning it’s an advanced driver assistance system rather than true autonomous driving.

Recent user experiences highlight this reality. Tesla drivers report that the system requires constant vigilance, with frequent warnings to pay attention and potential “strikes” that can lock users out of the system if they don’t maintain adequate attention. One user described being “completely shocked” when FSD made dangerous maneuvers during a road trip, leading them to cancel their subscription out of safety concerns.

Tesla’s own data shows the challenge ahead. The latest FSD version achieves approximately 500 miles between critical disengagements, while Tesla’s stated goal is to surpass human safety levels at 700,000 miles between accidents according to NHTSA data—a gap that remains substantial.

The Austin Launch: A Limited First Step

Tesla’s most significant autonomous driving milestone in 2025 is its planned June launch of “unsupervised” FSD in Austin, Texas. However, this represents a narrow implementation rather than the broad consumer rollout many have anticipated.

The Austin service will operate as a geo-fenced ride-hailing platform using Tesla’s internal fleet of vehicles. Key limitations include:

• Geographic constraints: Limited to specific areas within Austin
• Fleet-only operation: Available only through Tesla’s ride-hailing service, not for individual Tesla owners
• Remote supervision: Despite being “unsupervised,” Tesla will employ teleoperation staff to monitor the vehicles remotely
• Small scale: Starting with just 10-20 vehicles before potential expansion

This approach closely mirrors Waymo’s strategy, which Tesla CEO Elon Musk previously criticized. Waymo has been operating similar services for years, completing over 250,000 paid rides weekly across multiple cities.

The Gap Between Promise and Reality

Tesla’s autonomous driving program has consistently fallen short of Musk’s timelines. Since 2016, the company has promised full self-driving capabilities “next year” with remarkable consistency—and equally consistent delays.

Recent admissions from Tesla executives reveal the scope of the challenge. Ashok Elluswamy, Tesla’s head of AI and self-driving, acknowledged in a rare candid interview that Tesla is “lagging a couple years” behind Waymo in autonomous driving capabilities.

The Austin launch represents what critics call “moving the goalposts.” Rather than delivering the unsupervised self-driving in consumer vehicles promised since 2016, Tesla is launching a limited commercial service that bears little resemblance to what early FSD customers purchased.

Technical Challenges Remain

Tesla’s camera-only approach to autonomous driving continues to face significant hurdles. Unlike competitors who use multiple sensor types including LiDAR, Tesla relies exclusively on cameras and AI processing. While this approach offers potential cost advantages, it has yet to demonstrate the reliability needed for true autonomy.

Hardware complications add another layer of complexity. Tesla recently acknowledged that millions of vehicles with Hardware 3 (HW3) will not be capable of full self-driving, requiring costly upgrades for affected customers.

Looking Ahead: Gradual Progress

Despite the challenges, Tesla has made measurable progress. The company reports that customers have driven over 3 billion miles using FSD (Supervised), providing valuable training data for its neural networks. Tesla increased its AI training compute capability by 400% in 2024, demonstrating continued investment in the technology.

Plans for 2025 include:

• Expansion of the Austin robotaxi service if successful
• Potential launches in California and other U.S. markets
• Continued improvement of FSD capabilities for supervised use
• International rollout of FSD (Supervised) to Europe and China, pending regulatory approval

The Verdict

Tesla in 2025 remains tantalizingly close to, yet still meaningfully distant from, true full autonomy. The company’s supervised FSD represents sophisticated driver assistance technology, but falls short of the autonomous revolution Musk has long promised.

The upcoming Austin launch marks genuine progress toward commercial autonomous transportation, but its limited scope and reliance on remote supervision underscore the remaining technical and regulatory challenges. For Tesla owners hoping to summon their cars or sleep during highway drives, that future remains elusive.

What’s clear is that autonomous driving is proving more complex than early optimists predicted. While Tesla continues to push boundaries and may achieve limited commercial success in 2025, the era of widespread, unsupervised autonomous vehicles remains a work in progress—one that will likely extend well beyond this year.

The question isn’t whether Tesla will achieve full autonomy, but when—and whether the company’s approach will ultimately prove successful against competitors who have chosen different technological paths. For now, Tesla drivers must remain alert, hands on wheel, ready to take control at a moment’s notice.