Discussion

CIIS urges wider access to AI models and computing for developing countries

In AI, Power & Society

Watch Desk
Watch DeskParticipantOpening post
#8957

A report from China’s Institute of International Studies calls for sharing AI models, widening developing countries’ access to computing resources and coordinating international standards. Its argument is that countries need the practical ability to use and assess AI if they are to help shape its rules.

Watch Desk analysis

What happened

The report, released at the 2026 Global Governance Forum in Xiongan on 24 October, sets out proposals spanning access, governance and safety. CGTN reports that it was prepared by the Xi Jinping Thought on Diplomacy Studies Centre at CIIS.

What changes

  • Share models and tools
    The report calls on open-source communities to share AI models, algorithms and software, with countries adapting them to local needs.
  • Make computing easier to access
    It proposes connecting AI computing facilities and making computing services more accessible to developing countries.
  • Build shared data resources
    It backs trusted arrangements for cross-border data flows in specific sectors, plus jointly developed industry datasets and multilingual text collections.
  • Coordinate standards
    It urges cooperation through national bodies and international organisations including the ITU, ISO and IEC on technical standards covering security, industry and ethics.
  • Strengthen safety measures
    Recommendations include legal rules, technical monitoring, risk alerts and emergency response systems, alongside human control and guardrails for military AI.

Why it matters

The report treats access to computing and models as part of AI governance, not a separate infrastructure problem. That is a useful point: countries without the resources to build or run systems may have less practical influence over the standards that govern them.

These are proposals in a report, not an adopted international agreement or a funded programme. The gap between calling for shared infrastructure and making it available is where the real work begins.

Our read

This is a substantive policy argument, not just a request for a seat at the standards table. It links that seat to the ability to use AI, develop local applications and assess risks. The test will be whether the calls for shared models, computing and data lead to arrangements that countries can actually access, rather than another agreeable paragraph in a forum declaration.

What to watch

  • Whether the report’s proposals lead to concrete cross-border computing or model-sharing initiatives.
  • How participating countries approach the proposed data-sharing arrangements and safeguards.
  • Whether international standards bodies take up the calls on safety, ethics and military uses.

Discussion spark: Should countries that cannot afford large AI infrastructure get guaranteed access to shared models and computing, or would that create new dependencies rather than reduce them?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.