The International Telecommunication Union (ITU), the World Health Organization (WHO) and the World Intellectual Property Organization (WIPO) will convene the third meeting of the Global Initiative on AI for Health (GI-AI4H) in Hangzhou, China, from 16 to 18 September 2026.
The meeting will bring together policy-makers, regulators, health practitioners, researchers, technical experts, representatives of civil society and industry. It will focus on putting responsible AI for health into practice, strengthening the work of GI-AI4H and supporting cooperation in response to the needs of Member States.
GI-AI4H brings together the complementary expertise of ITU, WHO and WIPO in telecommunications and technical standards, health, and intellectual property. Its work encompasses standards and governance, collaboration and capacity-building, and the responsible use of AI in real-world health settings.
Meeting focus
Building on previous GI-AI4H meetings held in Riyadh, Saudi Arabia, in 2023 and Singapore in 2025, the third meeting will focus on the practical, responsible and sustainable implementation of AI for health.
The provisional programme is expected to include:
- visits showcasing AI for health research, development and implementation in China;
- a high-level policy dialogue on AI governance and cross-border collaboration;
- parallel technical sessions on priority areas in AI for health; and
- progress reports and consultations involving GI-AI4H Working Groups and Topic Groups.
Participants will exchange practical experience, review progress and identify opportunities for cooperation across different national and regional settings.
Registration
Registration is open to the public, subject to approval. Participants may choose to attend in person or online when registering. Please note that in-person participation is self-funded. Applicants will be notified by email whether their registration has been approved.
If needed, create an ITU user account before registering.
Opportunities to contribute
GI-AI4H invites contributions from the public through two submission opportunities.
1. Submit an implemented AI for health use case
GI-AI4H invites submissions of implemented AI for health use cases that offer practical lessons for health authorities, policy-makers, implementers, researchers and other stakeholders.
The call focuses on AI-enabled interventions that have already been deployed in a real-world health service or health system setting and that contribute directly to primary health care, universal health coverage or health system strengthening.
Submissions should briefly describe:
- the health system problem and implementation setting;
- the AI-enabled intervention;
- its deployment and scale;
- results and supporting evidence;
- governance arrangements;
- equity and accessibility considerations; and
- key implementation challenges and lessons.
GI-AI4H will seek to select submissions that reflect geographical balance across WHO regions. All selected submissions will be developed into brief case profiles for inclusion in the meeting report, subject to editorial review and applicable clearance processes. A subset of selected cases will be invited for presentation at the meeting.
Submission does not guarantee selection. Selection, presentation or inclusion in the meeting report does not constitute endorsement by ITU, WHO or WIPO of any technology, product, institution or approach.
Submission deadline: 23:59 CEST (UTC+2) on Thursday, 20 August 2026
2. Join the GI-AI4H Benchmarking Challenge
The Global Initiative on AI for Health invites participation in the GI-AI4H Benchmarking Challenge, an open global challenge through which health AI models can be evaluated using clinical data without disclosing the data to the model developer or the model to the institution holding the data.
The challenge tests privacy-preserving evaluation approaches that allow clinical data to remain under the control of the institutions that hold it while submitted models are evaluated securely.
The challenge is designed for evaluation rather than model training. Submitted models are evaluated in their existing form and are not trained during the challenge.
Participants may take part as solution providers contributing privacy-preserving evaluation methods, data hosts contributing clinical evaluation tasks and datasets, or model developers submitting models for evaluation. Solutions are assessed against three criteria: correctness, non-disclosure and cost. Models are scored for each clinical task using metrics defined by the institution contributing that task, with subgroup results reported where the data support such analysis.
Participation is free of charge and open to hospitals, universities, ministries of health, research institutions, companies and other interested organizations.
Interim results from the challenge will be reported at the third GI-AI4H meeting in Hangzhou, China, from 16 to 18 September 2026. The results do not constitute regulatory approval, certification or endorsement of any model.
Interim submission cutoff: 23:59 UTC on 28 August 2026
Submission window closes: 23:59 UTC on 18 December 2026
Final results: January 2027