In July 2026 the Ministry of Industry and Information Technology (MIIT) launched its 2026 open call for innovation tasks in the industry and information technology field. This batch lists artificial intelligence as a heading of its own, alongside the future-industries heading, each with several tracks beneath it.
The AI heading has four tracks: foundational technology, application-enabling technology, intelligent products and equipment, and common supporting technology. The future-industries heading has five: humanoid robots and embodied intelligence, atomic-scale manufacturing, high-performance separation and purification equipment for bio-manufacturing, brain-inspired intelligence, and frontier laser manufacturing.
What each of the four tracks points to
Foundational technology points to the model, framework and toolchain layer. For a company whose core assets are models or algorithms, this is the entry point closest to its own technology stack.
Application-enabling technology points to the ability to embed AI in specific industries and business processes. It stresses solving problems in real scenarios rather than demonstrating general capability.
Intelligent products and equipment points to software-hardware products — terminals, robots, specialised devices. This track connects directly to embodied intelligence and smart terminals.
Common supporting technology points to enabling capabilities such as datasets, evaluation, security and tooling. These are often the unglamorous but bottlenecked links in the chain, and the gaps that smaller teams can enter.
The logic of the “open call” mechanism
The open call is not the conventional “apply for a topic, then review and approve” process. It works as “publish the task, post it publicly, and those who can deliver take it up”. The task and expected targets are defined first by the competent authority, then claimed by companies or consortia. It embodies “whoever can do it, does it”, which is relatively friendly to companies with strong engineering capability but no traditional research-institute pedigree.
Paired with this is consortium application: many tasks cannot be completed alone and often require a lead entity to join with upstream and downstream partners. This aligns with the wider policy direction of promoting intelligent technical upgrading of software companies, discussed in our AI law hub.
The compliance groundwork applicants must prepare in parallel
Beyond the technical task, a national-level task also places demands on a company’s governance capability. In practice, applicants for open-call tasks usually have to explain three things in their submission: the degree of independent controllability of the technical solution, whether data sources and use are compliant, and whether they have the engineering organisation to deliver continuously.
Data is the most easily underestimated part. The sources, licensing chains and personal-information processing arrangements for training and evaluation datasets all need traceable explanations. For the boundaries of these duties, see our Algorithm Filing and Data Compliance service pages.
Intellectual property is the same. Open-call tasks often take “technical outcomes” as their deliverable; if patent layout, open-source use and ownership of results are not arranged in advance, later commercialisation and financing will be blocked. See our IP Protection service page.
Three points for AI companies
First, judge where you sit among the four tracks. Foundational-technology companies look at the first track, industry-solution companies at the second, terminal and robot makers at the third, and data and security tool providers at the fourth.
Second, build a consortium early. The open call encourages upstream-downstream collaboration; going it alone puts you at a disadvantage in review, and finding complementary technology and scenario partners early is more effective than assembling a team at the deadline.
Third, treat compliance materials as part of the application. Data-source explanations, IP inventories and ethics and security governance mechanisms will all be examined in review and during task execution. For support on application compliance materials, data and IP arrangements, see our Data Compliance and Commercial Contracts services.
This article is based on public information on the 2026 open call for innovation tasks in the industry and information technology field launched by the Ministry of Industry and Information Technology in July 2026. It is for reference only and does not constitute legal advice. Specific application requirements are subject to the official application guidelines.