In August 2025, Xiangdao Travel will release its AI "3+2" strategy; In March 2026, Didi launched the "Droplet" AI Travel Assistant v1.0; In April of the same year, T3 Travel submitted its prospectus to the Hong Kong Stock Exchange, listing "AI+Travel" as the second core strategy; In June, Cao Cao Chuxing announced the launch of a comprehensive AI transformation. In less than a year, all leading ride hailing platforms in China have incorporated AI into their strategic core.

This is not a simple technical trend. The ride hailing industry has gone through a decade of rapid growth, with traffic dividends peaking, subsidy wars failing, and excess supply of transportation capacity. The industry has entered a deep water zone of stock game. When the extensive model of "burning money for growth" comes to an end, platforms have to seek efficiency from technology. And AI, at this moment, provides an opportunity for full chain reconstruction from operational scheduling to user experience, from cost control to business models.

The deeper motivation lies in the valuation logic of the capital market. If a ride hailing platform is only defined as a "transportation intermediary", the price to earnings ceiling is clearly visible; But if labeled as an "AI driven technology platform", the valuation model jumps from the traditional transportation industry to the technology stock sequence. It is not a coincidence that the intensive capital actions of T3 Travel, Cao Cao Travel's transformation, and Xiangdao Travel's updated prospectus coincide highly with the timeline of AI strategy release.

AI is rewriting the competitive rules of the ride hailing industry. Whoever completes the transition from the "dispatching tool" to the "decision-making center" first may win the ticket to the next stage in this game.

AI Path Differentiation

After sorting out the AI layout of each platform, a clear context emerged: the Internet school represented by Didi and the car enterprise school represented by T3 travel, Cao Cao travel, and enjoy the road out behavior showed significant differences in their attitude, rhythm, and focus on AI due to different genes.

As the largest player in the industry, Didi's AI layout started earliest and has the widest coverage. In March 2026, Didi launched the AI travel assistant "Droplet" v1.0; In June, after upgrading to version 8.0, the app will embed AI into every business line. According to Didi's 2026 first quarter performance report, "Didi" relies on more than ten years of accumulated real operational data and supports more than 90 refined service tags such as "fresh air, large trunk, and smooth driving". In the field of security, Didi deeply integrates AI capabilities with its security system and introduces a human-machine collaboration model of "AI intelligent analysis+manual review by security experts".

Didi's AI strategy presents a "global penetration" feature, not limited to a single functional module, but embedding AI as a underlying capability into the entire chain of scheduling, security, customer service, and user experience. In May 2026, Didi and Zhipu jointly established an AI Exploration Laboratory; At the same time, Didi, as one of the first partners, will integrate its core ride hailing services into WeChat AI Agent. From building its own AI capabilities to opening up its travel skill interface, Didi is attempting to build itself as the "travel fulfillment base" of the AI era.

Different from Didi's starting from the Internet, T3 Travel, Caocao Travel and Xiangdao Travel rely on vehicle manufacturers, and naturally have the industry closed-loop thinking of "car building+travel". Their AI layout is more inclined to make strategic reserves for the era of autonomous driving.

T3 Mobility is the most systematic AI powered company among car manufacturers. In April of this year, T3 Travel officially submitted its prospectus to the Hong Kong Stock Exchange. According to the prospectus, T3 Mobility is positioned as a practitioner of the "AI+Mobility" model, utilizing AI to optimize demand forecasting, vehicle scheduling, and resource allocation. Its core foundation is the "Leading Path Model" jointly created with China Telecom, which has been implemented in three major scenarios: intelligent scheduling, safety assurance, and driver and passenger services: intelligent scheduling reduces the driver's empty driving rate by 17.5%; 37 AI digital employees independently undertake 85% of driver service cases; By 2025, the AI service takeover rate will reach 90%. In terms of financial data, T3 Travel's gross profit margin in 2025 has jumped from 0.4% in 2023 to 13.0%, achieving an operating profit of 7.44 million yuan, which is the most direct quantitative proof of AI efficiency improvement.

Cao Cao's AI transformation in transportation is the most thorough. On June 18th, Cao Cao Chuxing announced the launch of a comprehensive AI transformation, released the RoboX strategy, and proposed to build a "globally leading physical AI mobile technology platform". Cao Cao Travel has restructured its organizational structure around AI and established AI and RoboX business units. The core of RoboX's strategy is to focus on three major elements: intelligent customized cars, intelligent driving technology, and intelligent operations, and to lay out services such as Robotaxi and Robovan. It proposed the "Double One Hundred Thousand Plan": to deploy a total of 100000 Robotaxis and 100000 Robovans by 2030.

Cao Cao Chuxing stated in an interview with the Economic Observer that it has developed large-scale capabilities in transportation capacity scheduling, asset management, user services, and more. In particular, relying on the industrial ecology of Geely Holding Group, it will launch customized vehicle models specifically developed for travel scenarios in 2023, providing a foundation for the large-scale operation of Robotaxi.

Xiangdao Travel is one of the earliest car companies in the industry to propose a systematic AI strategy platform. In August 2025, Xiangdao Travel will release its AI "3+2" strategy, establishing key projects around intelligent business, intelligent research and development, intelligent brain and office, personal assistants, and other directions. The independently developed "Xiangdao Intelligent Brain" system uses end-to-end online reinforcement learning technology to reduce the distance of ride hailing services nationwide by 9.1% and increase drivers' hourly income by 6.8% year-on-year.

From a timeline perspective, Xiangdao Travel was the first to release a systematic AI strategy, after which T3 Travel capitalized its AI strategy, Didi completed productization, and Cao Cao Travel achieved a comprehensive strategic upgrade. Although the pace of each platform has its own sequence, the direction is highly consistent.

From the perspective of effectiveness, the AI efficiency improvement of T3 Travel has been supported by financial data, with a gross profit margin increase of over 12 percentage points in three years; Didi's "Little Droplets" have entered the stage of large-scale operation; The "smart brain" of Xiangdao Travel has quantifiable improvements in two indicators: driving distance and driver income. Cao Cao's comprehensive transformation of transportation is still in its early stages, but its goals are ambitious.

Driving Forces and Industry Changes

The collective embrace of AI by ride hailing platforms is a result of technological upgrades on the surface, but at a deeper level, it is the combined effect of industry structural contradictions and capital market logic.

The ride hailing industry has entered the stage of stock game. According to data from the Ministry of Transport, the total number of ride hailing orders in China in May 2026 was 977 million. Multiple cities have issued capacity saturation warnings, with the number of drivers continuing to increase but the unit price per kilometer continuing to decline. The previous model of relying on subsidies to grab users is no longer effective. Bai Wenxi, Vice Chairman of the China Enterprise Capital Alliance, pointed out that the growth rate of domestic ride hailing orders has slowed down, and the industry has entered stock competition. AI is the key to solving the problems of "idle capacity" and "supply-demand mismatch" - by optimizing the dispatching algorithm through large models, the empty driving rate can be reduced (currently the industry average is still over 30%), directly increasing profits.

The valuation reconstruction of the capital market is the most direct external driving force for AI transformation. Marketing industry expert Chen Rongrong believes that after price subsidies expire, AI precision services become the core barrier to retaining users, and "AI storytelling" is also the key to obtaining high valuations.

The valuation of ride hailing platforms in the capital market has been under long-term pressure. Traditional ride hailing services are low profit businesses, and capital has limited imagination about them. Bai Wenxi stated that incorporating "AI+travel" into the core strategy is to break away from the label of a simple "transportation intermediary" and transform into a technology company to obtain higher capital market valuations.

T3 Travel submitted its application in April, Xiangdao Travel updated its prospectus in May, and Cao Cao Travel announced its AI transformation in June. Their intensive capital actions are highly synchronized with the release of AI strategies. As Wang Guanqiao, a person in the Internet industry, said, if the platform only defines itself as "intermediary" or "service company", the price earnings ratio given by the capital market will be very low; But if "AI+Mobility" is included in the core strategy, the valuation model instantly jumps from traditional transportation industry to cutting-edge technology stocks.

The rigid constraints of compliance and safety cannot be ignored. The ride hailing industry is under strong regulation, and AI driven real-time facial recognition, travel deviation warning, and fatigue driving monitoring have become essential options for platforms to avoid policy risks. Bai Wenxi pointed out that under the background of strong regulation, these AI capabilities have changed from "bonus points" to "mandatory options". The safety model for T3 travel has reduced the number of traffic accidents per million to 15.7, far below the industry average.

The secondary awakening of data assets provides technical feasibility. The ride hailing platform generates a massive amount of orders, GPS tracks, in car audio, and road condition information every day. Wang Guanqiao pointed out that in the past, most of these data were "useless" or only used for simple heat map display; With the development of generative AI and multimodal big model technology, these spatiotemporal data have become excellent fuel for training vertical big models. Whoever can first transform data assets into AI capabilities can build industry barriers.

Taking seats ahead of schedule in the era of Robotaxi is a deep strategic consideration for ride hailing platforms to collectively embrace AI. The Ministry of Industry and Information Technology has released a draft for soliciting opinions on mandatory national standards for L4 level autonomous driving. Several cities across the country have opened up the commercial charging operation qualification for Robotaxi without safety officers. Robotaxi is considered the ultimate form of ride hailing, and once autonomous driving matures, labor costs will approach zero.

Chen Rongrong predicts that in the next 3 to 5 years, a hybrid transportation network with both manned and unmanned driving will be formed, and AI will dynamically allocate transportation capacity based on cost, efficiency, and scenarios. Wang Guanqiao further pointed out that the current AI transformation is essentially training future "hybrid operation systems", allowing platforms to seamlessly switch between manned and unmanned hybrid operations during the transition period. Bai Wenxi also believes that the current AI layout is actually paving the way for L4 level Robotaxi. Whoever masters the complex real road condition AI decision data holds the ticket to future travel.

Cao Cao expressed a similar judgment in his response to the Economic Observer: AI is moving from the digital world to the real world, and transportation services such as travel and freight are becoming important infrastructure in the AI era. Therefore, Cao Cao chose to create a physical AI mobile technology platform and build an intelligent transportation system that connects digital intelligence and the physical world.

Efficiency improvement and pattern reshaping

The impact of AI on the ride hailing industry is becoming apparent on multiple levels. The improvement of operational efficiency is the most direct and measurable. T3's leading road model reduces the driver's empty driving rate by 17.5%, and 37 AI digital employees independently handle 85% of driver service cases; The "smart brain" of Xiangdao Travel has reduced the pick-up distance by 9.1% and increased the driver's hourly income by 6.8%. Behind these numbers is a substantial decrease in operating costs and a systematic improvement in service efficiency.

At the user experience level, ride hailing apps are transforming from cold forms of "entering point A to point B" to intelligent assistants that can understand vague instructions. Didi's "Droplet" supports more than 90 refined service tags such as fresh air, large trunk, and smooth driving, transforming users' personalized preferences into platform executable matching conditions. Wang Guanqiao pointed out that in the future, AI assistants can actively remind users to call a car based on their calendar, weather, and habits, and even recognize drunk passengers, pet users, and other scenarios through multimodal perception, and adjust services accordingly.

From a business model perspective, deep restructuring is happening. In the short term, AI will push ride hailing from being "labor-intensive" to being "algorithm intensive"; In the medium term, both manned and unmanned driving coexist, and AI dynamically allocates transportation capacity based on cost, efficiency, and scenario; In the long run, if autonomous driving becomes mature, travel prices may decrease by more than 50%.

But the coin has another side. Bai Wenxi reminds that high R&D investment in AI will further squeeze the survival space of small and medium-sized platforms, and the Matthew effect in the industry will intensify. Chen Rongrong also pointed out that after the maturity of autonomous driving, the driver group will shrink, but at the same time, it will give rise to new professions such as AI trainers, remote safety officers, and vehicle maintenance.

The ride hailing industry is undergoing a qualitative change from informatization to intelligence. In traditional mode, AI is an auxiliary tool that helps dispatch orders and plan routes; In the native mode of AI, AI has been upgraded to a decision-making center, with everything from vehicle scheduling and dynamic pricing to safety monitoring and service quality control being led by AI. This shift means that the core elements of industry competition have shifted from capacity scale and subsidy intensity to algorithm accuracy and data thickness.

The industry landscape will also accelerate differentiation, with one end being the global AI operation platform represented by Didi, which builds ecological barriers with its scale and data advantages; On the other hand, platforms such as Cao Cao and T3, which rely on the background of car companies, are betting on Robotaxi with their closed-loop capabilities of "customized cars+intelligent driving+operation". Small and medium-sized platforms caught in the middle will face greater survival pressure.

The commercialization process of Robotaxi will be a key variable determining the industry's ultimate outcome. According to a research report by Dongwu Securities, it is expected that the number of Robotaxis in China will reach about 520000 by 2030, with a shared travel fleet penetration rate of about 10%. Whoever can lead the commercialization loop in this competition may occupy the core position of the future travel market. From this perspective, AI is no longer an optional bonus for ride hailing services, but rather the infrastructure that determines the next stage of life and death.

(Author Wang Shuaiguo)