Think Piece Author:Dr. Qiang (Braven) Zhang,Professor of Public Policy at the School of Government, Beijing Normal University (BNU), Director of Token for Good Hub (T4G), and a Board Member of the China Volunteer Service Federation (CVSF). 

Two milestones fall within the same twelve months, yet almost nobody has connected them. In December 2025, the UN opened the International Year of Volunteers for Sustainable Development, anchored by a flagship report, the 2026 State of the World’s Volunteerism Report, showing that 2.1 billion people – 34.5 percent of the world’s working-age population – volunteer each month. In July 2026, the UN also convened the first Global Dialogue on AI Governance in Geneva, where 193 member states were told that artificial intelligence capabilities are outpacing our ability to govern them. 

Volunteering is the planet’s largest distributed human infrastructure. The question is not whether AI will reshape it, but whether volunteering – humanity’s oldest form of unpriced cooperation – can shape AI back. 

My argument is threefold. AI’s real contribution to volunteering is coordination, not substitution. Its greatest risk is turning measurement into surveillance. Its decisive test is whether it narrows or widens the gap between the countries that build AI and those that live with its consequences. 

1. The bottleneck is coordination, not compassion

The 2026 report contains a quietly radical finding: direct, person-to-person volunteering (25 percent of working-age people) is more than twice as common as organisation-based volunteering (11.7 percent). Goodwill is abundant; what is scarce is the infrastructure to route it accurately and deliver it when it matters. 

That is precisely what AI excels at. In Keqiao District, Shaoxing, a mother whose child was in crisis at midnight posted a request on the ‘Good Neighbour’ platform. A resident in the same community with a counselling certificate was matched and responded within minutes. In Chongqing’s Jiulongpo District, an AI application builds skill profiles for volunteers and ‘demand lists’ for neighbourhoods, closing the loop from need

identification to matching, delivery, feedback, and recalibration. The UN Volunteers program reached the same conclusion from the other direction: its AURA platform, launched in 2026, uses multilingual AI to match candidates to assignments while keeping ‘a human-in-the-loop at all stages’. 

The design principle worth naming is ‘volunteer-in-the-loop’: machines route, rank, and remember; people decide, care, and take responsibility. 

2. Measure what matters – and stop the measurement from being gamed

Measurement is the organising theme of the 2026 report, whose Global Index of Volunteer Engagement (GIVE) assesses individual and community benefits, economic value, and the enabling environment – the latter explicitly including data and technology. AI can serve that agenda well: tools that turn scattered volunteer feedback into evidence, models that detect crisis signals from satellite and local-language data, and assistants that ease the reporting burden for overstretched organizations. During China’s 2026 ‘99 Giving Day’, an AI assistant built for the philanthropic sector offered 42 task-specific capabilities – drafting proposals, processing receipts, and writing field reports – enabling staff to spend less time on paperwork and more on people.

But China also offers a cautionary case. When volunteer hours became a credential used in school and workplace evaluations, a grey market emerged: hours for sale, proxy check-ins, and impersonated accounts. Between June and August 2026, China’s cyberspace and social-work authorities ran a nationwide campaign against exactly this. The lesson generalises: a measurement system that optimises logged hours will simply automate the distortion. Volunteer data must flow first to volunteers and communities, then to auditors. 

3. From digital divide to civic dividend

Here, the global stakes are clearest. Frontier AI compute is highly concentrated – roughly three-quarters of the capacity of the world’s 500 most powerful supercomputers is in the United States, and about 15 percent in China, leaving 191 of 193 states unable to independently audit the systems they are asked to govern. In Geneva, the Secretary-

General proposed a Global Fund for AI to help low-income countries build capacity rather than dependence. 

Volunteering is the natural channel: already local, already trusted, already present where state services are thin. China’s own ‘AI+’ action plan, issued by the State Council in August 2025, targets over 70 percent adoption of smart terminals and AI agents by 2027 and makes ‘AI + global cooperation’ one of its six pillars, committing to treat AI as an international public good. That commitment should now be operationalised: a Global South network of shared, locally tuned, open models for needs-matching in disaster response and community health; volunteer-in-the-loop as a normative default; and a common measurement aligned with GIVE, so data remains comparable while remaining locally owned. 

Volunteering has survived every previous technology cycle because it has never been about efficiency. It is about people choosing to show up for strangers. AI earns its place on one criterion: does it help more people show up – and does it ensure that, when they do, the world finally counts them?