Audrey Tang

Audrey Tang is a Taiwanese civic hacker and g0v co-initiator (see Person entry) and the most internationally-recognised voice associated with Taiwan's grassroots civic-tech community, with the Polis-based "broad-listening" deliberative-democracy infrastructure the corpus tracks through the vTaiwan and g0v platforms, and with the participatory-AI-governance register that runs from the 2014 Sunflower Movement through the 2024 Alignment Assemblies of AI. She is tracked here as a Voice because her public output — the co-authored Plurality: The Future of Collaborative Technology and Democracy (2024, with Glen Weyl and the Plurality Community), international talks on digital democracy and participatory AI governance, the "humour over rumour" pandemic-era meme-communications model, and the conceptual framings of "broad listening" and "uncommon ground" — has done more than any other single public-output corpus to install Taiwan's civic-tech experience into global movement-building vocabulary around participatory AI governance.

The Plurality framework

Tang's most substantial public-output vehicle is Plurality: The Future of Collaborative Technology and Democracy (2024), co-authored with economist and technologist Glen Weyl and the Plurality Community. Written collaboratively on GitHub, published into the public domain under CC0, community-translated into over a dozen languages, and available free online, the book itself embodies the recursive-public model it describes. Its core argument navigates between "the extremes of techno-libertarianism and centralized AI governance", proposing a third path in which digital technology enables "collaboration across social differences" — collaborative diversity rather than consensus aggregation or centralised AI judgment. For the make-AI-good movement, the Plurality framework is the most substantial single published account of what participatory AI governance should be building toward rather than only what it should resist, and the 2025 Right Livelihood Award — citation: "For advancing the social use of digital technology to empower citizens, renew democracy and heal divides" — recognises Tang's work as the cumulative carrier of this register through the g0v, vTaiwan, and Plurality arc.

"Broad listening" and vTaiwan

The most propagated methodological concept in Tang's public-output register is "broad listening" — the Polis-based deliberative-democracy approach that reveals "areas of convergence rather than amplify extremes" by mapping participants' positions into an opinion landscape and surfacing where groups agree across a large statement set. The crucial design choice is removing the reply button, preventing the adversarial thread dynamic that makes online political deliberation collapse into polarisation. Tang's account of why this matters appears in the 2018 MIT Technology Review profile: removing the reply button "drastically reduces the motivation for trolls to troll" so that public servants come to see commenters "not protesters or mobs, but actually people with distinct expertise". The Right Livelihood citation names this as elevating "bridging ideas over loudest voices" and uncovering "uncommon ground" — points of agreement between ideologically different people — as the deliberative output.

Tang carries the broad-listening method forward into her AI-governance register in the June 2024 Reboot Democracy interview: the March 2024 Alignment Assemblies of AI, launched through the Ministry of Digital Affairs in partnership with the Collective Intelligence Project, Anthropic, OpenAI, The GovLab, and the GETTING-Plurality research network, applied the same broad-listening infrastructure to invite "everyday citizens to co-govern AI in the context of information integrity". The Alignment Assemblies represent the downstream application of the vTaiwan-era method to AI governance questions — Tang's voice connecting the 2014-era civic-tech infrastructure to the 2024-era AI-deliberation field.

Participatory AI governance register

Tang's most recent and actively-propagating register is participatory AI governance — the proposition that building AI for public benefit requires co-design with the public rather than delivery to the public. The Reboot Democracy interview states the position directly: "progress can only be achieved when AI is grounded in participation: to build AI for the people, with the people." The register refuses the two dominant AI-governance framings — technocratic expert deliberation and market-mediated consumer choice — and grounds AI governance in the broad-listening civic-tech tradition Tang's voice anchors.

The participatory-AI-governance register distributes through Tang's post-cabinet portfolio as Cyber Ambassador Taiwan, Oxford Institute for Ethics in AI Senior Accelerator Fellow, TED 2026 Guest Curator, and advisory roles at the Mozilla Foundation, Project Liberty Institute, Ethereum Foundation Plurality Initiative, and DemocracyNext International Advisory Council — a constellation of distributed institutional vehicles for the single participatory-AI-governance register rather than a single institutional home.

"Humour over rumour" and the pandemic-era communications model

During Taiwan's COVID-19 response, Tang and the g0v community developed the communications approach the Right Livelihood citation names as "humour over rumour": a pre-bunking strategy using government memes to anticipate falsehoods before they spread, ensuring "everyone could see the same facts at the same time, so panic didn't spread faster than the virus." The Conversation's documentation of the operational format names it the "2-2-2" method — a government meme response in 20 minutes, 200 words, 2 images — with notable examples including viral dog-meme campaigns. Alongside the meme strategy, civic-tech community members built the Mask Maps: open-data applications showing real-time pharmacy mask inventory across Taiwan, which The Conversation records under Tang's framing of "reverse procurement" — a bottom-up, community-built alternative to top-down government distribution. The structural enabler, per the same account, was "two-way trust between government, civic hackers, and citizens" developed from open-source software community practices — the same foundation the g0v community and the vTaiwan deliberative-democracy infrastructure rested on.

Signature framings

Four framings recur across Tang's public output and have done the most to install her register into the broader make-AI-good movement.

• "Broad listening." The Polis-based method that removes the reply button to surface bridging positions and uncover "uncommon ground". The framing carries through the vTaiwan documentation, the Alignment Assemblies of AI, and the Right Livelihood citation as the named deliberative-democracy method.
• "To build AI for the people, with the people." The participatory-AI-governance register from the Reboot Democracy interview. The framing carries through Tang's post-cabinet advisory portfolio as the distributing vector for the civic-tech approach to AI deliberation.
• "Humour over rumour." The pandemic-era pre-bunking communications model — government memes in 20 minutes as faster counters to disinformation than corrections or bans. The framing carries through civic-tech pandemic-response and digital-resilience advocacy circles internationally as a named methodology reference.
• "Fast, fair and fun." Tang's standing description, named in the Right Livelihood citation, of what democracy must be: responsive to needs, inclusive of all voices, and energising rather than exhausting participation. The framing operates as the through-line from the g0v civic-hacker ethic to the Plurality framework.

Why this is a Voice entry

A Voice entry is created here, rather than additional structure on the Person entry, because Tang's public-facing output is itself the load-bearing object the corpus needs to track: the Plurality framework as the most substantial single published account of participatory AI governance's positive vision; the "broad-listening" conceptual framing and its Polis-based vTaiwan-era infrastructure; the participatory-AI-governance register — "to build AI for the people, with the people" — and its application through the 2024 Alignment Assemblies of AI; and the "humour over rumour" and "reverse procurement" communications models that carry the civic-tech approach to disinformation and crisis response into international advocacy circles. The corpus's voices slice had no Taiwan-anchored voice, no civic-tech-and-participatory-governance voice, and no voice carrying the Polis-based broad-listening method into the AI-governance register — this entry closes all three gaps. Affiliation and biographical detail are recorded on the linked Person entry per the corpus's Person/Voice split.