On March 17, Alibaba announced what it called the world’s first enterprise-grade AI native work platform, Wukong, positioning the product as a turnkey way for companies to deploy always-on AI agents across teams. Wukong will be available as a standalone application and will be embedded into DingTalk, Alibaba’s enterprise collaboration suite, with invitation-only testing beginning immediately. The platform promises seamless connection to enterprise DingTalk accounts, secure access controls and integration with corporate applications, lowering the technical and administrative friction that typically constrains AI roll-outs inside companies.
The launch is less a technological breakthrough than a distribution play: by wiring an agent platform into DingTalk’s mass of corporate users, Alibaba is trying to shortcut adoption and push AI from labs into everyday business workflows. Enterprises can expect agents that automate routine tasks, surface contextualised information and act on behalf of teams, rather than generic chatbots. For Alibaba this offers a route to capture higher-value cloud and application revenues while shaping how Chinese firms standardise AI agents in HR, sales, operations and customer service.
Also on March 17, Didi introduced Xiaodi v1.0, an AI travel assistant that translates user requests into executable service labels — more than 90 currently — such as “spacious trunk” or “smooth driving” and handles complex scenarios like escorting elderly passengers or business pick-ups. Xiaodi combines these preference tags with live traffic, timing and location data to suggest the “best” matching vehicle or routing plan, supports one-click bookings and can recommend transfer solutions for long journeys. The feature reframes competition in ride-hailing from pure speed and availability to precision of match: not just getting a car quickly, but getting the right car for the trip.
These two product debuts sit alongside several other signals of China's AI commercialisation cycle. Weimeng Group reported 2025 AI-related revenues of 116 million yuan and stated that the company achieved adjusted net profit and the first positive operating cash flow since listing — evidence that SaaS vendors are finding ways to monetise AI. Meanwhile, XR brand VITURE, together with NVIDIA and Stanford’s Cong Lab, showcased an XR-AI lab automation solution at NVIDIA’s GTC, illustrating the push to combine immersive hardware, GPU compute and AI for specialised scientific workflows.
Taken together, the announcements illustrate a shift in Chinese tech strategy: from headline-grabbing models and futuristic promises to pragmatism — products that generate measurable value, slot into existing customer bases and can be monetised. Alibaba’s strategy seeks to capitalise on an installed enterprise platform, Didi is mining behavioral and rating data to improve user certainty, and smaller SaaS players are showing the early payoffs of embedding AI into paid offerings. High-performance partnerships, such as VITURE’s with NVIDIA and academia, show that capital-intensive, niche AI applications are also advancing.
The moves do not come without risk. Integrating agents deeply into enterprise systems raises questions about data governance, access controls and regulatory oversight, especially as China tightens standards for AI services. Competition will be intense: Tencent, Baidu, Huawei and cloud specialists are likewise racing to offer enterprise AI foundations, and multinational cloud providers remain active in adjacent markets. Success will depend not only on technical capability but on trust, ecosystem lock-in and the ability to turn agent interactions into sustainable revenue streams.
For international observers, the most consequential takeaway is that AI in China is entering a phase of commercial consolidation. The emphasis on plug-and-play enterprise tooling and contextualised consumer assistants signals that the next wave of winners will be those who translate model performance into operational efficiency, regulatory compliance and repeatable monetisation.