MK and Pony.ai Sign Memorandum on Autonomous Driving Technology in Japan

MK and Pony.ai signed a memorandum of understanding on Oct. 8, 2026, as strategic partners to exchange views on the development of autonomous driving technology in Japan. The signing took place at Kyoto Smart City Expo 2026, held at Keihanna Plaza. The memorandum covers discussions on developments in autonomous driving technology in Japan, safety and operational challenges, regulatory and permitting requirements, and the potential introduction of robotaxi services.

When taxi operators such as MK, which operates in nine cities in Japan, are involved, robotaxis enter a phase in which the operating entity, regional rollout and permits must all be addressed together, not just technical testing. Pony.ai cites operations in nine countries and more than 100 million kilometers of cumulative autonomous driving mileage, but deployment in Japan still requires permits under relevant laws and regulations. Commercialization also depends on other conditions.

What is new in this announcement is that MK, a major domestic taxi operator, has established a framework for discussions with China-based Pony.ai on autonomous driving technology in the Japanese market. The memorandum, however, does not call for the immediate introduction of robotaxis. It is limited to exchanging views on development trends, safety, operational challenges, and regulations and permits. In other words, the companies are at the stage of working through the institutional and operational design that precedes commercialization. While the partnership’s scope is presented broadly, the conditions for actual operation remain limited at this point. In a related article published on Oct. 18, 2026, [Toyota rapidly closes the gap with Tesla and Chinese rivals in autonomous driving, and moves seriously into robots and flying cars, Nikkei xTECH reports](/en/news/toyota-chases-tesla-and-chinese-rivals-in-autonomous-driving-nikkei-xt-716397), we covered Toyota’s plans for mass production of autonomous vehicles and robot development using field data. The MK-Pony.ai deal can be read not only as a move in vehicle and AI development, but also as an effort to secure a taxi operator’s capacity to support actual services in Japan. Conversely, it underscores that robotaxis cannot be launched without an operating entity, regional networks and the ability to handle permits, even when the technology provider is strong. From an industry perspective, the partnership could broaden the competitive focus in autonomous driving beyond algorithms and vehicle performance to include the ability to put services into operation locally. MK has a taxi network operating in nine cities nationwide, while Pony.ai cites operations in nine countries and more than 100 million kilometers of cumulative driving experience. The two companies’ strengths therefore lie in different areas. The memorandum, however, only calls for examining the possibility of deployment; it does not specify which cities or vehicle models would be involved or what operating model would be used. Going forward, it will be necessary to assess permit requirements, the transition from static display to on-road operation, and safety evaluations tailored to Japan’s traffic environment.

Pony.ai said it operates in nine countries and has logged more than 100 million kilometers of cumulative autonomous driving mileage. MK operates taxi services in nine cities nationwide. The companies said they would deepen their understanding of the safety and operational challenges of autonomous driving technology in Japan’s traffic environment. The introduction of robotaxi services would require permits under relevant laws and regulations, and the companies said they would consider future initiatives separately based on the outcome of their discussions. The companies exhibited together at the venue, displaying Pony.ai’s seventh-generation robotaxi outdoors without operating it. MK’s press release also said Pony.ai uses “Virtual Driver,” its full-stack autonomous driving technology integrating proprietary software, hardware and services, and is working to build a business model that enables mass production and deployment independent of vehicle model.

In Japan, operators with networks spanning multiple cities, such as MK, could provide a base for robotaxi services, while deployment will depend on permits and operational planning in each region. The focus of autonomous driving discussions is not only vehicles and AI, but also how the dispatch, fleet management and local-response capabilities of existing taxi operators can be integrated.