On 24 November 2024 Beijing E-Town, formally the Beijing Economic-Technological Development Area, revised and issued a package of policies to build an AI-native industry innovation highland, issued as Jing Ji Guan Fa [2024] No. 29. Two numbers carry most of the signal: around 100 core companies across the AI chain, and industry scale above RMB 100 billion by 2026.
“AI-native” is the operative phrase
Most district-level AI policies promise subsidies to companies that settle in. This document goes further by describing the district itself as AI-native, meaning the industrial base — manufacturing, logistics, healthcare, government services — is to be reorganised around AI rather than merely hosting AI tenants. That framing matters legally as much as commercially, because it converts open questions about data access, compute allocation and model deployment into procurement and cooperation arrangements that companies can actually contract for. The district’s earlier positioning is set out in our note on the Beijing E-Town AI city plan.
How it sits with the municipal and national layer
The district package is a downstream instrument. It is best read together with the city-level Beijing AI industrialisation plan and the Beijing software and AI application plan, which supply the qualification criteria and application scenarios the district then funds. Nationally, the same logic of “AI plus industry” appears in the State Council AI Plus action opinion. E-Town is not an outlier; it is one of several cities competing on the same playbook, as shown by the Hangzhou AI innovation highland plan.
What AI companies should actually do
First, decide whether the commercial value justifies the compliance cost of relocating or registering a subsidiary in the zone. Incentives are usually paid against local registration, local tax contribution and locally recognised R&D expenditure, which means the paperwork is the deal.
Second, treat compute and data access as the scarce items. Where a policy bundles computing power or public data resources, that is often worth more than cash, and it depends on the district’s national computing network service platform connections.
Third, sequence the applications. Qualification is typically staged, and an early filing for a pilot scenario is easier than a retrospective subsidy claim after the project has already been built.
One caution is worth stating plainly. An incentive package of this kind is a commitment about support, not a promise of a market. Before moving a team or a subsidiary, it is worth checking whether the customers the district expects to provide actually exist within reach, and whether the cooperation agreements on offer are drafted with enforceable delivery terms rather than expressions of intent.
This article is based on Beijing E-Town’s policy document Jing Ji Guan Fa [2024] No. 29 and public information and is for reference only; it does not constitute legal advice.