Where to next? Insights from autonomous-vehicle experts

A 2025 survey by the McKinsey Center for Future Mobility indicates that autonomous vehicle adoption timelines have slipped by one to two years, with large-scale robo-taxi deployment now projected for 2030. While the United States and China lead the market with over 700,000 weekly autonomous rides combined, experts are increasingly viewing Level 3 automation as a niche premium product rather than a mass-market standard. This shift reflects growing concerns over high development costs and the technical complexity of moving beyond Level 2+ driver-assistance systems.

The autonomous vehicle (AV) sector is experiencing a global ramp-up, with Level 4 robo-taxis now delivering more than 700,000 paid rides per week across major markets. According to data from McKinsey, the United States leads with 450,000 weekly rides—driven largely by Alphabet’s Waymo—while China follows with 250,000. Despite these milestones, a January 2025 survey of 91 industry leaders reveals that most deployment timelines have been delayed by 12 to 24 months. Large-scale robo-taxi rollouts are now expected in 2030, while the viability of fully autonomous trucking and private L4 urban pilots has been pushed back to 2032.

Regional progress remains uneven, as experts predict the United States and China will maintain a three-to-seven-year lead over Europe and the rest of Asia. This gap is attributed to more agile commercial structures, superior regulatory support, and urban environments better suited for AV testing compared to dense European city centers. Additionally, these leading regions benefit from larger market sizes and a more robust AI and software talent pool. Consequently, while Europe has hosted over 35 pilots, it continues to trail the aggressive commercialization seen in North America and China.

The strategic focus for privately owned vehicles is also shifting, with 49% of experts now predicting that Level 2+ (L2+) functions will dominate the mass market through 2035. This is a reversal from 2023, when a majority of respondents expected Level 3 (L3) or higher systems to be the standard. The pivot is largely due to the high costs of development and validation, alongside a lack of funding that has turned L3 functionality into a niche, premium-tier offering. Furthermore, despite advancements in AI, the estimated costs for achieving higher levels of autonomy have actually increased as the industry grapples with the true complexity of full deployment.