連線上都不行:RightsCon 2026 取消事件的臺灣第一人稱紀錄

A first-person record from a Taiwanese session host at RightsCon 2026, and steward of Lezismore, an anonymity-first queer digital community in Taiwan.

When RightsCon 2026 was forced not only out of Lusaka but out of its online venue as well, I was one of the Taiwanese civil society participants whose exclusion had reportedly become a condition for the summit to proceed. The exclusion was not only about travel. According to Access Now, the conditions conveyed informally through multiple sources included excluding Taiwanese participants from online participation as well.1

That detail is the reason I am writing.

The facts, briefly. On April 27, one day after a Zambian government press release endorsed RightsCon, Access Now received a phone call from Zambia’s Ministry of Technology and Science. Diplomats from the People’s Republic of China were pressuring the Government of Zambia because Taiwanese civil society participants were planning to attend in person. The conditions for resumption, as Access Now later described them, would have required moderating specific topics and excluding “communities at risk, including our Taiwanese participants, from in-person and online participation.” Access Now refused. Rather than comply, Access Now let the summit not proceed.2

That refusal is the only reason this account is possible. Without it, no version of these conditions would be on the public record. I want to name that clearly before going further: Access Now drew a red line that made first-person accounts like this one possible at all.

I am the steward of Lezismore, an independent, self-hosted queer community in Taiwan that has been running since 2015 on minimal data, behavior-based accountability, and what I have elsewhere called “good friction” as an alternative to identity gates.3 We have no government funding, no venture capital, no NGO backing. We are also not visible in the way platforms are usually counted, because the architecture that would make us legible to outside trackers, scrapers, and aggregators is the same architecture we have built Lezismore to refuse. That refusal is a design choice, and the condition under which a queer public of this kind can keep existing at all.

Last year I spoke at RightsCon 2025 about my own platform’s work. This year my role was different: I had been preparing, with an Australian co-host, a roundtable session on identity verification, freedom of expression, community-level moderation, and the rising barriers to digital participation that cut across all of these. The format was deliberately built to be a working conversation among policymakers and other practitioners, with the agenda structured around shared questions rather than centered on any single platform. The session was about how participation in a digital civic space can be designed without using identity as the gate, a question that matters for queer counterpublics and for many other communities whose members face risk when made legible. The conditions reportedly imposed for the summit’s resumption asked the convening institution to install exactly such a gate, drawn along national lines.

I do not know whether I was specifically targeted. I do not believe I need to. The conditions reportedly demanded would have excluded participants in exactly my position: Taiwanese civil society participants working at intersections of marginalized communities, anonymity infrastructure, and platform governance. That is what the public record can support. That is what I will work from.

I want to sit with the online clause for a moment, because that is where this case becomes a digital rights problem rather than only a travel-and-conferencing one.

RightsCon is, by design, a hybrid summit. Online participation is one of the primary mechanisms by which the global digital rights community includes people who, for reasons of cost, visa regimes, surveillance risk, disability, or geographic distance, cannot or should not travel. It is also the modality through which the people most directly subject to digital harm can take part in the conversations the field is having about them.

The form of exclusion at issue here is less familiar to the digital rights field than the politics behind it. In the field’s most familiar vocabulary, exclusion of a specific national group from a digital civic space is imagined as a network-layer intervention: cables disrupted, bandwidth throttled, IPs blocked. The Access Now-led #KeepItOn coalition has spent nearly a decade documenting that pattern as the internet shutdown.4 What happened at RightsCon sits at a different layer: political pressure converged on the institution that controlled access to the venue.

This is not new for Taiwanese civil society. For decades, Taiwanese participation in international convenings has been quietly negotiable: sometimes survived, sometimes traded away in the calculations diplomatic hosts make to keep their events going. What was different this time is that Access Now refused to make that trade. The condition was named, the refusal was named, and the public record of both now exists. That is rare, and it is rare partly because, in many other instances over many other decades, the path of least resistance has been to exclude us quietly and proceed.

The reason this case matters beyond Taiwan is that the same mechanism, pressure applied through diplomatic and administrative channels and economic pressure rather than network controls, is now available wherever there are gatekeepers and there are external states with reach. Taiwan has been the long-running test case for this kind of exclusion. Many other communities are now within range of it.

I want to be careful about why this position, mine, matters at all in this account.

What Lezismore is, in the language I have come to use for it, is a queer counterpublic. By that I mean a queer public organized in deliberate refusal of extraction. It has chosen invisibility to outside trackers, scrapers, and aggregators because the alternative, being legible to anyone who wants to compile, sell, or weaponize what is said inside, would close down the kind of community life it was built to protect. That choice is part of a wider design pattern counterpublics adopt when the attention-economy logic that runs most platforms doesn’t work for them. Lezismore’s architecture is built around what I have elsewhere called “good friction”: deliberate choices about onboarding, pace, and what the platform makes easy or holds back, so that risk is absorbed at the system level rather than passed to users. The community trades the metrics of visibility maximization, the frictionless onboarding, ambient discovery, and algorithmic reach, for something else: the conditions under which a sex-positive queer Taiwanese public can keep speaking, disagreeing, taking risks together, and continuing to expand on its own terms.

This logic shows up in many other publics. Trans communities adapting to the U.S. policy environment, sex workers building infrastructure outside platforms that have deplatformed them, dissidents under regional surveillance, indigenous communities whose knowledge sovereignty depends on selective sharing: in different ways and on terms specific to each, many of these communities have made architectural decisions that overlap with Lezismore’s. They survive, each on their own terms, by refusing the visibility-equals-legitimacy bargain that mainstream platforms treat as default.

Counterpublics of this kind do not show up in the metrics by which platforms are usually evaluated. That is part of how they work. It is also part of why convenings like RightsCon matter to them disproportionately. The spaces where the people sustaining these communities can speak with peers, name their own experience, and contribute to how the digital rights field understands its own subjects, are few. When a third state can collapse one of those spaces through diplomatic pressure, the loss extends well beyond a single conference. What is being squeezed is the infrastructure through which counterpublics make themselves audible to each other and to the field.

This is one account from one position. The harm of this cancellation reaches a great deal further. Zambian and African digital rights communities lost a long-planned global convening, the first RightsCon to be held in Southern Africa, and the first to return to the African continent since RightsCon Tunis in 2019. The cancellation played out partly through their own state institutions. Their accounts of what happened, and of what was already at stake before the Taiwan question entered the picture, are the accounts I most hope to read next, and the ones I think this piece should sit alongside.

A note, finally, on why I am writing this at all.

For ten years, I have written sparingly about the constraints that shape work like Lezismore’s. Articulating those constraints publicly tends, in my experience, to expand rather than narrow the surface area of attack. Quiet has been a learned response to that structural condition. Speaking publicly has cost, including cost to the people and community my speech reaches. Before drafting this piece, I worked through those implications with care, and made a deliberate judgment that this specific case warrants the risk. That judgment is specific to this moment. It does not generalize.

Access Now has documented what happened. Human Rights Watch, Index on Censorship, Article 19, and others have made statements.5 What was missing, until now, was a first-person account from one of the people the exclusion was designed to silence.

This is that account.

The next question for the digital rights field goes beyond how to respond to this cancellation. It is what to do with the fact that the convening infrastructures the field depends on, and the counterpublics that depend on those convenings, can be unmade by a sequence of administrative and diplomatic pressure converging on the institutions that host them. The work is in the conditions that let counterpublics survive when the political ground shifts under them: the resources, the hosting arrangements, the tooling, the legal protection, and the governance flexibility to adapt without compromising on what made these communities possible in the first place. Some of those communities will look like Lezismore. Many will not. What they share is a logic the field has not always been ready to recognize: resilience for counterpublics often runs through deliberate selective invisibility, through choosing pace over reach, through governing in ways that mainstream metrics will register as failure.

Writing cannot, by itself, resist exclusion. Documentation can keep exclusion from pretending it never happened. The accounts that come next, from the people who were not supposed to be in the room, are what will make this record useful.

This is not a conclusion. It is a record. Records are where accountability begins.


  1. Access Now, “A statement to our community about why RightsCon 2026 will not take place in Zambia” (May 1, 2026): https://www.rightscon.org/rc26-statement/
  2. Tech Policy Press, “RightsCon Canceled After Zambia Requires ‘Full Alignment’ With ‘National Values'” (April 30, 2026): https://www.techpolicy.press/rightscon-canceled-after-zambia-requires-full-alignment-with-national-values/.
    WIRED, “The Chinese Government Pressured Zambia to Cancel the World’s Largest Digital Rights Conference” (May 1, 2026): https://www.wired.com/story/the-chinese-government-pressured-zambia-to-cancel-the-worlds-largest-digital-rights-conference/.
    Human Rights Watch, “Zambia: Summit on Human Rights, Technology Effectively Canceled” (May 1, 2026): https://www.hrw.org/news/2026/05/01/zambia-summit-on-human-rights-technology-effectively-canceled
  3. Electronic Frontier Foundation, “Speaking Freely: Shin Yang” (March 2026): https://www.eff.org/pages/speaking-freely-shin-yang
  4. Access Now, “#KeepItOn: fighting internet shutdowns around the world”: https://www.accessnow.org/campaign/keepiton/
  5. ARTICLE 19, “Zambia: RightsCon cancellation is a blow to freedom of expression”: https://www.article19.org/resources/zambia-rightscon-cancellation-is-a-blow-to-freedom-of-expression/.
    Index on Censorship, “Zambia censors an international conference…on censorship” (April 2026): https://www.indexoncensorship.org/2026/04/zambia-censors-an-international-conferenceon-censorship/.

When It Happened

I was buried under more than two hundred post-accident medical documents, along with a mountain of cases, research files, work, and to-dos. In that haze, when I received an email from Access Now—the host organization of RightsCon, the global summit on human rights in the digital age—titled “Important update: RightsCon 2026 will not proceed in Zambia”, I spent a few minutes confirming whether it was a phishing attempt.

It wasn’t.

The reason? Not stated, but the situation seemed urgent—participants were advised not to travel to Zambia. We were less than a week from opening day; just days earlier I had still been going over the rundown and tech setup with my co-host.

I immediately reached out to my international co-host. No response, no read receipt. I worried they were already on a plane. Other international scholars and friends scheduled to attend or support the session—some received the news literally on their way to the airport and had to turn back. Everyone was thrown into chaos. Nobody knew exactly what was going on.

A few days earlier, news had broken that President Lai Ching-te had been “denied transit” through the Czech Republic and Germany. So in that moment, a thought genuinely flickered through my mind: “Okay—maybe it’s not directed at Taiwanese people. Maybe the whole event is being shut down. Everyone is affected.” That thought briefly comforted me. I have never enjoyed being in the spotlight—anywhere, in any context. So the idea that it wasn’t “just me” or “because of me” made me feel a little more at ease.

Mom! I Got the Presidential Treatment!

Then, just a few hours ago—around midnight on May 2 Taiwan time—Access Now sent another email. This time, the full official statement.

On April 27, one day after a government press release endorsed RightsCon, we received a phone call from MoTS about an urgent issue and were told that diplomats from the People’s Republic of China (PRC) were putting pressure on the Government of Zambia because Taiwanese civil society participants were planning to join us in person.

What the government wanted from us in order to lift the postponement was conveyed to us informally from multiple sources: in order for RightsCon to continue, we would have to moderate specific topics and exclude communities at risk, including our Taiwanese participants, from in-person and online participation.

Holy shit, this really is highly targeted at Taiwanese people. Damn.
So I really did get presidential treatment, huh.

Mom! Look, your daughter just got the presidential treatment! Hahahahahaha—someone take a photo, post it to my story!!!!! (I did not.)

(I really did not post a story.)

(Calm down.)

Given how many times in my life I have been excluded, ejected, or silenced, I’ve grown used to it to the point of laughing it off. It’s just that this time it has escalated to the level of international certification—like I had somehow entered the Boss Level, thrill mixed with terror.

After laughing, I read the email several times.

Reading those lines again, I saw three layers of exclusion: (1) the ‘specific topics’ required to be moderated, (2) the ‘at-risk communities’ required to be excluded, and finally (3) the explicitly named Taiwanese participants—including online participation.

When I was a kid, we used to play Bingo. The first person to fill in all the conditions and form a winning line was the grand champion. I never won, but god, I always envied that thrill of victory. To finally experience it like this—well, thank you, everyone. Thank you, teachers. Thank you, fellow students. (wipes away tear)

Anyway—I realized I was sitting at the multiple intersection of this exclusion list.

The intersection looks like this: A Taiwanese person, born and raised. A sexual minority. The steward of a sex-positive, anonymity-first independent community.

For more than ten years, the daily work of someone in this position has not been the kind of thing you can list on LinkedIn, the kind of thing you can get authority figures to endorse, or get well-known people to recommend.

It is the work of judging possible risks behind tens of thousands of anonymous, plain-text submissions—submissions with no background information and no personal data attached.
It is evaluating an account that keeps changing IP addresses, while its behavioral pattern remains unmistakably suspicious; acting carries risk, and not acting carries risk, too.
It is designing layers of governance so that anonymity and safety do not cancel each other out, while preserving as much user autonomy as possible.
It is bearing the reality that no matter what governance decision you make, someone will be angry, hurt, or unsatisfied.
It’s researching and thinking through why a given governance decision works in some communities and contexts but fails in others.
It’s, with no external resources, no funding, no institutional or organizational backing, no structural safety net of any kind (or rather: where one wrong move could see the system destroy you), proving—every single day, every single day—that this community deserves to continue existing tomorrow.

This kind of work is silent.

It is infrastructure work. The defining feature of infrastructure is that it is meant to be like air, like sunlight, like utilities or the internet—when functioning, it should feel natural, unnoticed. When it fails, it explodes in the most public place possible, dealing devastating damage to the community it serves.

Over ten years, it has not, in any meaningful sense, exploded.
To prevent that explosion, I paid in silence.

This is where the multiple intersection sits.

The Cost of “Being Seen”

Over the past decade, “community management” has become a discipline everyone wants in on—platform follower growth, traffic monetization, personal branding. Trust has been collapsed into visibility, quantified into metrics, until this has become almost the only recognized model of human exchange in our era.

I do not buy this model. Or at least, I do not accept that it has been hyped up and engineered into the only option available to contemporary human beings. This is not a fashionable critique I just decided to make today, nor a take I generated last week. More than ten years ago, I was already betting my life, ongoing, against this model. That is how I became the kind of steward people describe, too easily, as ‘low-profile’.

I gradually grew out of the angry-young-person mode of trying to explain everything. Eventually, my answer became: “I’m super introverted.” “I just don’t like the spotlight.” “I’m the low-profile type.” This answer effectively pauses anyone who wasn’t really that curious to begin with. It’s a precisely calculated simplification—lightly papering over the structural problems actually at work underneath.

Those structural problems layer on top of each other:

First, the position of being a sexual-minority independent community itself.
In Taiwan, when what you run is an explicitly sex-positive, anonymity-first community that does not rely on mainstream commercial models, you are never standing on the same line as the mainstream “pretty, visible, palatable LGBTQ+ stories.” The higher the visibility, the higher the risk of being misread, co-opted, or attacked.

Second, the instability of the legal framework and its enforcement.
Across adult content, privacy, platform liability, and the protection of minors, Taiwanese law has never been designed for an independent community of this scale—and the framework warps and contracts every few years. To survive while carrying tens of thousands of community members through that unstable framework, the steward has to constantly evaluate the potential consequences of every action. The more “high-profile” a steward with a face, an image, a personal brand becomes, the more every gesture, every word, directly amplifies all those potential consequences.

Third, governance sustainability.
For an independent community without government subsidy, without VC investment, without NGO backing, the most precious resource is not even attention—it is time. Going viral once and then collapsing or shutting down and going home to bed is, in the modern era, entirely unremarkable. It can even help a personal brand—example: “List the projects you’ve worked on.” But the long-standing needs of sexual minorities and marginal communities have very few Taiwanese, independent, non-Big-Tech-dependent infrastructures capable of carrying them long-term. Choosing the latter means walking slowly for ten years, in silence.

Fourth, the flattening specifications of the attention economy.
To be seen by the mainstream, you have to enter the mainstream’s narrative framework. There is nothing wrong with this; it is normal. You always have to compress yourself into a certain shape—apply the perfect foundation, draw eyeliner that photographs well, so that you can be solidified into a beautiful image, tagged with a hashtag that reproduces easily. Different fields have different demands: mainstream society has its voyeurism, its patronizing pats on the head for sexual minorities; lesbian communities have their own internal anger and expectations; tech founders are asked for scale, growth, and optimized user experience; researchers are asked for academic ethics, formats, procedures. None of these demands are neutral. All of them require you to actively become a version of yourself they can compress, comprehend, consume, extract from, exchange for something else.

Stack those four things on top of each other, and “being low-profile” stops being a personal preference of mine. To some real extent, it is a necessary condition for surviving in this position, this work, this decade.

And even if you set aside those structural conditions—even if I were willing to be more visible—”being seen” carries its own price.

Over the past ten years, I have been fortunate to receive some interview requests, collaboration invitations, media inquiries. The common thread among most of them: I am asked to enter a predetermined narrative framework.

Show your face. That’s the baseline. Show your face, give them a face, give them a clip-worthy, screenshot-worthy visual hook for reels and reposts.

Then come the content demands: “What are some hot girl-on-girl stories on the platform?” “We’d really like to record a podcast where the focus is your personal story, your sexual preferences, your sexual practices.” “What lines do lesbians most often use to hook up?” “How do lesbians have sex? Help us teach the public properly!” “So do trans people have XXX?”

These questions are not all malicious. They are the reflexes the attention economy trains. Under the logic of traffic, everyone has been trained to know what gets clicked, shared, fought over until everyone picks a side. The collaborators inviting me genuinely believe they are “helping me get exposure”—they sincerely think “giving you the opportunity to be exposed” is a form of reciprocity.

But for me, this is a complete misunderstanding of what a sexual-minority independent community even is. Once you accept that framework, you have already chosen to let your governance principles concede ground. As an individual, of course not being seen makes me feel the ache at times. But if the price of being seen is flattening the entire complexity of the community—disrespecting its history and origins, ignoring the risks and difficulties it faces—just to be exchanged for “being seen” by whom, exactly? Seen, and then what? In the face of vague, abstract “exposure,” I have to seriously consider what happens if a sudden wave of strangers crashes into the infrastructure and what risks that creates for the community.

So I’d politely say, “No. Could you adjust the framing? We can talk again.” And then the market mechanism of the attention economy would immediately classify me as “not worth collaborating with,” “not interesting enough,” “doesn’t know how to do marketing,” “too difficult to work with.” Some of these came from mainstream LGBTQ+ advocacy organizations themselves—because they are also struggling, also need to “be seen by the public.”

So I fully accept the verdict of being difficult.

In the logic of the attention economy, that is exactly what I am. This is a conscious, autonomous choice.

The Two-Sided Cost of Silence

On the other hand, silence is not a “self-righteous choice” either. It is, more often, because I already tried speaking, and the result was worse.

Many times. Every time.

Difficulty gets weaponized. The community has had, over the long term, malicious users who repeatedly probe the mechanisms and try to break things. As long as we hold the line on anonymity, on not collecting personal data, on not running public trials—those forms of malice will not disappear.

If that’s hard to grasp, think about the kinds of weaknesses and costs people commonly attribute to democratic freedoms. That gets you part of the way there.

Total silence: you are erased. Loud public speech: you get weaponized.

Ten years has been spent, every day, carefully maintaining the balance between those two. Lose that balance once, and I may not be the only one who pays; the cost may fall on another community member’s life. Hold steady, and what you’ve earned is only the breathing room of the next day. There are no pats-on-the-back. There is no reassuring solution. Being told “you worked hard” does not pay for dinner.

You cannot please both sides. You ultimately have to choose. And bear that choice.

Even before writing this piece, I spent several hours doing the necessary security review. This was not impulsive, and it was not a choice that pushes the community or the people close to me into unnecessary risk.

On “Digital Inclusion”: Where Two Kinds of Invisibility Meet

What is most worth writing down about this event is not the simple conclusion that “Taiwanese people were excluded.” It is that the event made two kinds of “unseenness” collide on a single list.

One is the invisibility I have chosen and paid for over the past decade—the price one pays for refusing the market mechanisms of the attention economy.

The other is the invisibility of Taiwan itself as a subject—name-changes imposed in international organizations, ruptured diplomatic relations, transit denials, and now even civil society participation in international venues becoming a bargaining chip.

These two look very different. One is the logic of the attention economy and personal branding; the other is the logic of geopolitics and diplomatic capital. But the underlying logic is fundamentally similar: deciding who gets to appear, who must disappear, and on what conditions appearance is permitted.

According to Access Now’s statement, what was conveyed informally through multiple sources was that, for RightsCon to continue, specific topics would have to be moderated and at-risk communities—including Taiwanese participants—would have to be excluded from both in-person and online participation. I do not know whether the actors involved knew the full scope of the work I do, and I have no way to determine whether I was a specific target of this action.

But this event stacked these two layers onto each other—and stacked them on this position: a sexual-minority independent community steward + a Taiwanese host of one session at an international digital governance summit + a high-risk subject area + platform governance research. The intersection of those four things contains very few people. That made me realize: I am sitting at the multiple intersection of these conditions.

RightsCon, at its core, is a venue for examining global civil society’s inclusion in the digital age. When the condition for the event continuing is “even the online participation of Taiwanese people, at-risk communities, and certain topics must be blocked”—then we get to see what digital inclusion actually looks like, with the kind of raw, bloodied clarity nobody planned for: the complex sum of “who qualifies,” “under what conditions,” “what forms of participation they are allowed to use, and what forms are denied.”

Only this time, those questions were answered by a single diplomatic phone call and a closed-door government meeting—leaving thousands of stunned international organizers, scholars, policymakers, and community practitioners outside the door, outside the borders.

In the End

Stopping everything else to write this immediately, after receiving the news, was because I sensed a rare opening.

For years I have stayed as quiet as possible—a strategic choice woven from regulation, platform governance, the specificity of the issue, privacy, and other interlocking considerations. But this time, if even those who have been excluded stay silent, the narrative imposed by others hardens into history.

Writing cannot directly resist exclusion. But writing can force exclusion to stop pretending nothing happened.

To be unseen can be a strategy: a choice, a refusal to sacrifice one’s complexity just to be seen blindly and incorrectly.

To be unseen can also be a fate someone else decides for you.

For the past ten years, my work and Lezismore’s work have lived in the former. This time, I am writing down the latter as well—and a part of my own story.

Between being seen and being unseen, you, too, can have your own choice.

來自一位 RightsCon 2026 臺灣議程主持者,也是以匿名的性少數獨立數位社群「濡沫」站長的第一人稱紀錄。

當 RightsCon 2026 不只被迫離開盧薩卡(Lusaka),連線上參與也無法成行,我是其中一位臺灣公民社會參與者;據報導,「將我們排除」是數位人權大會繼續舉辦的交換條件之一。這場排除不僅限於現場參與者。根據 Access Now,透過多方非正式管道傳達的條件,中國(中華人民共和國)施壓的內容,包括將臺灣參與者排除在線上參與之外。1

關於發生了什麼事,在此簡短說明:4 月 27 日,在尚比亞(Zambia)政府發布支持 RightsCon 的新聞稿之後一天,Access Now 接到尚比亞科技與科學部的電話。中華人民共和國的外交官正向尚比亞政府施壓,因為臺灣公民社會參與者計劃親自出席。Access Now 後來描述,恢復舉辦的條件,包括要求他們調整特定議題,並排除「有風險的社群,包含我們的臺灣參與者,於實體與線上參與之外」。Access Now 拒絕了。大會主辦方沒有妥協、接受條件,而是選擇了停辦。2

正是這份拒絕,讓此刻的紀錄得以存在。如果沒有主辦單位的拒絕,這些條件不會以任何形式進入公開紀錄。在繼續之前,我想清楚表示:Access Now 劃下了一條紅線,讓像這篇這樣的第一人稱紀錄成為可能。

我是濡沫站長。濡沫是臺灣一個獨立、自架站的酷兒社群,自 2015 年運作至今,運行原則包括最低限度的資料蒐集、以行為為基礎的問責,以及好的摩擦設計,作為身分驗證閘門的替代方案。3 我們沒有政府資金,沒有創投,也沒有 NGO 背書。因此,我們不會出現在主流平台用來展示數字、計算規模的指標競逐場合之中,因為濡沫從一開始,就刻意拒絕建構各種「被窺伺、被意淫、被去脈絡和便於傳播」的,那套由爬蟲、追蹤、大型平台集體狂歡所建設的架構。這個拒絕是有深刻痛感的、有意識的設計選擇,也是這種「酷兒公共」(queer public) 得以持續存在的條件。

去年我曾在 RightsCon 2025 分享過自己平台的工作。今年我的角色不同:我與一位澳洲共同主持者一起,原本正在準備一場圓桌討論,主題是身分驗證、言論自由、社群層級的內容審核,以及貫穿這些議題、不斷升高的數位參與門檻。這個圓桌的格式刻意設計為政策制定者與實務工作者之間的工作對話,議程圍繞共享的結構性問題展開,而非聚焦於任何單一平台。這場圓桌的核心問題是:如何設計一個數位公民空間,讓參與不必以身分作為通行的閘門。這個問題不只限於酷兒社群,更對「對抗性公共社群」(counterpublic) 也有意義,特別是那些「被識別」時就會承擔風險的社群。諷刺的是,雖然就在我們規劃這個圓桌討論的同時,中共施壓並提出恢復舉辦大會的條件,正是要求主辦機構裝設另一種身份閘門,而這次的身份閘門不是性別、不是年齡,而以國籍「臺灣」。

我無從得知自己是否被特定針對,也不認為我需要去深究。
接下來,我會從公開報導所說的三個交集繼續開展:在邊緣社群、匿名基礎設施、平台治理這些交集處工作的臺灣公民社會參與者(包括現場參與,以及線上參與者)。這是公開紀錄能夠支撐的論述。我會從這裡開展。

首先,我想把「線上」拉出來談。
因為這個地方,正是讓整個事件,從一個關於「旅行與會議安排」、「臺灣人又被打壓」的問題,真正升級成全球數位人權嚴峻問題的關鍵。

RightsCon (數位人權大會)本身在設計成混合式參與,包括線上與現場。線上參與是全球數位人權社群納入特定人群的主要機制之一,這些人因為費用、簽證制度、監控風險、身心障礙、地理距離等理由,沒辦法、或者不能前往實體會場。線上同時也是一個基本條件:讓最直接承受數位傷害的人,能夠在場參與這個領域對他們的處境所進行的討論。

在數位人權領域,一般說到封鎖或者干預,通常比較熟悉的形式和說法,會是網路層級的干擾:透過實體線路被中斷、頻寬被限縮、IP 被封鎖(還有臺灣海底電纜各種被切斷),試圖(或者成功)將某個特定群體或國家排除在數位資源之外,斬斷通訊可能,進而達到控制與孤立的目的。Access Now 主導的 #KeepItOn ,將近十年來都在持續記錄國際上各種網際網路封鎖模式。4 而這次在 RightsCon 發生的事,位於另一個層級:政治壓力聚攏到掌握會場進出的主辦單位身上。

對臺灣公民社會而言,這並不是新鮮事。幾十年來,臺灣在國際會議裡的參與一直是談判桌上的籌碼。有時候我們熬過去了,有時候則在主辦方為了讓會議繼續而做的取捨之中被犧牲。
這次不同的是:Access Now 拒絕做這種交易,而且公開留下清楚的紀錄。這很罕見。
而它之所以罕見,部分是因為過去幾十年來、在許多其他類似的場合,最省力的方式,一直都是把「我們」悄悄排除掉,然後繼續舉辦。

這個案例不只對臺灣重要。當一個制度、威權,可以透過對場域守門人施壓,不只透過物理手段,也能透過外交、行政、經濟手段實施,這個巨大的威脅,意味著對其他人而言,這些場域恐怕也不再安全。歷史上,臺灣長期以來都承受著這樣的壓迫。而今天,這種威脅的影響範圍,顯然不只有臺灣。

接下來,我必須認真說明:我的位置與這次紀錄的重要性。

我用來描述濡沫的詞彙,是「酷兒對抗性公共社群」(queer counterpublic)。意思是:這是一個以「拒絕被擷取」為前提而建構的酷兒。它不依附主流社交平台,拒絕爬蟲、拒絕規模化與大量個資搜集、身份驗證,是因為另一種主流選擇(讓任何想窺淫、消費、武器化社群內部私密對話的人,都能輕易讀取、帶走並利用)會直接終結這個社群被建造起來所要保護的共同生活。
這個選擇,是「對抗性公共社群」在主流平台「注意力經濟」邏輯失靈時所採用的更廣設計模式之一。濡沫的架構圍繞著我在另外一篇訪談中稱為「好的摩擦」(good friction)的設計:在註冊流程、節奏、平台讓什麼變得容易或有意拖慢上做出刻意的選擇,目的是讓風險被系統層級吸收,而不是轉嫁到使用者身上。
在主流平台「注意力經濟」邏輯下,濡沫作為「對抗性公共社群」,主動放棄了讓能見度最大化的數據指標,換取的是另一件事:讓一個情慾友善的臺灣酷兒對抗式公共社群裡的人們,能夠持續摸索、持續觀察與反思、持續發展自己的「怪」、持續承擔風險、繼續以各自的世界觀沉澱與拓展。

這個邏輯也出現在許多其他對抗性公共社群身上。在當前美國政策環境下調整生存策略的跨性別社群,在主流平台將其下架後另外搭建基礎設施的性工作者,在區域監控下行動的異議者,或者有知識主權分享區隔的原住民社群:以各自的方式、依各自的處境與特定條件,努力存活下來,甚至在不可見處蓬勃發展。濡沫與他們類似之處,在於都拒絕主流平台預設的「能見度等於正當性」的邏輯。

這類對抗性公共社群,並不會出現在一般評估平台時所用的指標裡,不會在你滑IG或者脆時自動推播給你。因為這種隱蔽性,這正是它們至關重要的運作方式之一。
因此,這也是為什麼像 RightsCon 這樣的會議,對它們的意義會格外重大。這類社群的維護者能與同儕對話、能命名自己的經驗、能為這個領域如何理解它的主體做出貢獻的場域,這種機會本來就不多。

當第三方國家能夠用外交手段輕易擊潰其中一個這樣的場域時,損失遠遠超過一場會議本身。被擠壓的,是讓對抗性公共社群能彼此聽見、也能讓其他領域聽見它們的基礎設施。

這是一份來自我個人單一位置的紀錄。
這次取消造成的傷害,遠遠不只如此。尚比亞與整個非洲數位人權社群,失去了一場籌備已久的全球會議:這原本是首次在南部非洲舉辦的 RightsCon,也是 2019 年突尼斯(Tunis)會議之後,首度重返非洲大陸。而取消的決定是透過他們自己的國家機構下達。
這些現場的人們,到底發生了什麼事?以及在臺灣議題進入視野之前,當地的人們已經背負什麼沈重的代價?這些才是我接下來最希望讀到的紀錄,也是這篇文章應該並列陪伴、而非取代的紀錄。

最後補充:為什麼我選擇寫這篇文章。

過去十年,我很少公開寫出濡沫社群的困境與運作限制。根據我的經驗,把困境和限制命名得越清晰,實務上反而會擴大被攻擊的可能性。沈默,是我面對這種結構性條件,從慘痛經驗裡所習得的暫時性回應。

公開發聲是有代價的,這個代價不只我自己承擔,也包括社群會被波及的部分。在動筆撰寫這篇文章之前,我謹慎評估過這些影響,最後做出審慎判斷:這次特定的案例,值得承擔這個風險。然而,這個判斷只屬於這個當下,不能自動類推到其他情況。

Access Now 已經留下了公開記錄。Human Rights Watch、Index on Censorship、Article 19 以及其他組織也都已經發表聲明。5 目前還匱乏的,是一份第一人稱的紀錄——來自這次被排除、被噤聲的那群人。
這就是那份紀錄。

數位人權領域接下來的問題,絕不只是「如何發表聲明回應這次取消」的範圍。更深的問題,其實這個事實:數位人權領域所依賴的會議基礎設施,以及需要這些會議才得以維繫的對抗性公共社群,能夠被一連串行政與外交壓力的累積所擊垮——只要這些壓力集中在主辦這些會議的機構身上,就能做到。
真正要做的工作藏在那些條件裡:那些讓對抗性公共社群在政治與權力的版圖位移時,仍能生存下去的條件——資源、主機安排、工具、法律保護,以及一種治理彈性,能夠在不犧牲讓這些社群最初得以存在的根本前提下,靈活調整。
其中一些社群會長得像濡沫。很多不會。它們共同擁有的,是一種這個領域或許尚未準備好接受的邏輯:對「對抗性公共社群」而言,韌性往往來自刻意的選擇性不可見,來自選擇節奏而非規模化觸及——一種會被主流指標直接判定為失敗的治理方式。

書寫本身無法直接抵抗排除。但寫下來,會讓排除無法粉飾、假裝它從未發生。接下來要出現的、來自那些其他被排除者的聲音、這些紀錄的集體呈現,才會讓我單一立場的這份紀錄真正發揮用處。

這不是一個的結論,而是一份紀錄。
紀錄,是問責的起點。


  1. Access Now, “A statement to our community about why RightsCon 2026 will not take place in Zambia”(2026 年 4 月 30 日):https://www.rightscon.org/rc26-statement/
  2. Tech Policy Press, “RightsCon Canceled After Zambia Requires ‘Full Alignment’ With ‘National Values'”(2026 年 4 月 30 日):https://www.techpolicy.press/rightscon-canceled-after-zambia-requires-full-alignment-with-national-values/
    WIRED, “The Chinese Government Pressured Zambia to Cancel the World’s Largest Digital Rights Conference”(2026 年 5 月):https://www.wired.com/story/the-chinese-government-pressured-zambia-to-cancel-the-worlds-largest-digital-rights-conference/

    Human Rights Watch, “Zambia: Summit on Human Rights, Technology Effectively Canceled”(2026 年 5 月 1 日):https://www.hrw.org/news/2026/05/01/zambia-summit-on-human-rights-technology-effectively-canceled
  3. Electronic Frontier Foundation, “Speaking Freely: Shin Yang”(2026 年 3 月):https://www.eff.org/pages/speaking-freely-shin-yang
  4. Access Now, “#KeepItOn: fighting internet shutdowns around the world”:https://www.accessnow.org/campaign/keepiton/
  5. ARTICLE 19, “Zambia: RightsCon cancellation is a blow to freedom of expression”:https://www.article19.org/resources/zambia-rightscon-cancellation-is-a-blow-to-freedom-of-expression/
    Index on Censorship, “Zambia censors an international conference…on censorship”(2026 年 4 月):https://www.indexoncensorship.org/2026/04/zambia-censors-an-international-conferenceon-censorship/ 

事情發生的時候

我埋在兩百多份的車禍後醫療文件,還有像山一樣的各種案件、研究資料、工作與待辦事項裡。也因為這樣的忙亂,當我收到全球數位人權大會 RightsCon 的主辦單位 AccessNow 來信,標題是 Important update: RightsCon 2026 will not proceed in Zambia。(重要更新:在尚比亞的2026全球數位人權大會停辦)我還花了幾分鐘確認這到底是不是詐騙。

是真的。
原因?沒有說明原因,但看起來情況緊急,建議所有參與者不要前往尚比亞。距離正式開始不到一週,我前幾天還在跟合作主持討論流程與技術細節。

我立刻聯絡了跟我一起合作的國外主持人,沒有消息,未讀未回。我擔心對方已經上飛機了。還有其他會到現場參與、支援的國際學者、朋友,有人是正要前往機場的途中收到,臨時取消行程,大家都被搞得烏煙瘴氣。沒有人知道確切到底怎麼回事。

由於幾天前知道賴總統經過捷克、德國「被拒絕過境」,當下我心裡確實閃過一絲「好吧,不是針對臺灣人,是整個會議。大家都無法進行」的想法,這讓我稍微鬆了口氣。我一直都是一個不喜歡站在聚光燈下的人,無論到什麼地方都是這樣。所以事件發生不是「只有我」或者「因為我」,讓我感覺比較安心。

媽!我也有總統級待遇啦!

直到幾小時前,臺灣時間 5/2 凌晨,主辦方 AccessNow 再次來信。這次是正式聲明全文。

On April 27, one day after a government press release endorsed RightsCon, we received a phone call from MoTS about an urgent issue and were told that diplomats from the People’s Republic of China (PRC) were putting pressure on the Government of Zambia because Taiwanese civil society participants were planning to join us in person.

(重點翻譯:4/27,主辦方接到尚比亞政府電話。中華人民共和國外交人員因臺灣公民社會參與者預計現場參與,向尚比亞政府施壓。)

What the government wanted from us in order to lift the postponement was conveyed to us informally from multiple sources: in order for RightsCon to continue, we would have to moderate specific topics and exclude communities at risk, including our Taiwanese participants, from in-person and online participation.

(重點翻譯:主辦方後來從多個非正式管道得知,若要讓 RightsCon 繼續舉辦,他們必須『調整』特定議題,並將有風險的社群排除於現場與線上參與之外;其中明確包括臺灣參與者。)

靠,還真的是高度針對臺灣人欸!幹!
哇靠我還真得到總統級待遇啊!
媽!快看你女兒得到總統級待遇啦哇哈哈哈哈哈哈~拍照打卡發限動!!!!!!(並沒有)

(我真的沒有發限動)
(冷靜)

由於人生中被排除、驅逐、消聲的經驗簡直不能太多,我早已習慣到會笑出來。
只是這次上升到國際認證層級,有種進入Boss Level的興奮感與恐怖感(?)

笑完之後,我把信讀了好幾遍。

三個條件之中,有三種被點名的對象,依照不同層次列出來:(1) 被要求調整的「特定議題」、(2) 被要求排除的「風險社群」,最後,(3) 被指名要求排除在外的「臺灣人——含線上參與」。

小時候大家會玩賓果遊戲,第一個集齊所有條件,還剛好成功連成一直線的人就是最大贏家。
我從來沒贏過,但是那種勝利的刺激跟興奮,真是讓我羨慕得要命。
現在居然是用這種方式體驗到,謝謝大家、謝謝師長、謝謝各位同學(拭淚)。

言歸正傳,
我意識到自己正在這份「排除名單」的多重交集上。

那個多重交集,是這樣的:
一個土生土長的台灣人。一個性少數。一個情慾友善、匿名為主的獨立社群的管理者。

過去十多年,這個位置的人,每天在做的事情,沒辦法條列到 LinkedIn 上展示,沒辦法找權威者背書、找知名人士做推薦。

是設計多種不同的治理機制,讓匿名性跟安全性不至於互相抵消,但又盡可能保留最大程度的使用者自主空間。
是評估一個不斷更換 IP 跳板但是行為模式明顯可疑的帳號,處理或不處理都有風險,該怎麼取捨。
是判斷上萬篇完全沒有任何背景資訊、個人資料的純文字匿名投稿背後的可能風險。
是承擔無論做什麼治理決策,背後總有人憤怒、受傷、不滿的現實。
是研究與思考,為什麼某個治理決策會在某些社群和情境中有效、在另一些情況下卻失敗。
是在沒有外部資源、沒有資金、沒有機構或單位背書,也沒有任何制度性後盾(正確來說,稍有不慎就會被制度摧毀)的情況下,每一天、每一天,重新證明這個社群值得明天繼續存在。

這份工作是無聲的。
這是一份基礎設施的工作。基礎設施的特點是,它像空氣、陽光,水電瓦斯網路,它運作妥當時,應該是自然、無感的。當它失敗的時候,會在最公開的地方爆炸開,直接對社群造成毀滅性的打擊。

過去十年,它沒有真正意義上的爆炸過。
為了不要爆炸,我選擇用無聲去交換。

這就是這個多重交集所在的位置。

「被看見」的代價

「社群經營」在近十年算是一種顯學,平台漲粉、流量變現、個人品牌經營,把信任等同於可見度,並且將之量化成數據,幾乎成了當代人與人之間,唯一被認可的交流模型。

我是不認同這種模型的,至少不認同它竟被吹捧、設計成當代人類唯一的選擇。
不是今天才跟風批判兩句、生一篇AI文,而是十多年前,我就用人生持續去賭這一份不認同。所以我成了一個「好像很低調的站長」。

從拼命解釋的憤青模式中慢慢學乖,後來,我的答案變成了「我超 I」、「天生就不喜歡聚光燈」「我就低調」。這個答案可以成功讓對方暫停其實也沒很在意的加減問問。
這是一個精確計算過的簡化答案,輕描淡寫地藏住背後真正在運作的結構性問題。

那是多層次的結構條件相互疊加:

第一,性少數獨立社群這個位置本身。
在臺灣,當你經營的是一個明確情慾友善、匿名為主、不靠主流商業模式的社群,你跟主流的「漂亮、可見、可被接受的同志故事」之間,從來不站在同一條線上。可見度越高,被誤讀、被收編、被攻擊的風險就越高。

第二,法律框架與實際執法的不穩定。
在成人內容、隱私、平台責任、未成年保護這些議題之間,臺灣的法律從來不是為這種規模的獨立社群設計的,且這份框架每隔一段時間就會變形擠壓一次。
背著數萬人的社群,要在這份不穩定的框架中存活,治理者必須時時刻刻評估每一個動作的潛在後果——當有臉、有形象、有個人品牌的治理者越「高調」,一舉一動、一言一行會直接放大所有潛在後果。

第三,治理的永續性。
一個沒有政府補助、沒有資本投資、沒有 NGO 撐腰的獨立社群,最寶貴的資源甚至不是注意力,是時間。爆紅一次然後垮掉、收掉,回家洗洗睡,在現代是毫不稀奇的常態。那對個人品牌甚至有幫助,ex:「列出你曾做過的專案」。
但是對性少數社群與邊緣議題長期以來的需求,卻沒有幾個臺灣自己的、獨立的、不依賴科技巨頭的基礎設施能持續承接。

第四,注意力經濟的扁平化規格。
要被主流看見,得進入主流的敘事框架。沒有對錯,這是正常。你總要把自己壓扁成某種樣子,上服貼的底妝、畫上相的眼線,才能被凝固進美麗的照片裡,被打上好複製的hashtag。
不同場域有不同的要求,主流社會對性少數社群的窺伺與摸頭,女同志社群內部自己的憤怒和期待;科技創業者也有量體、規模化、體驗最佳化的要求,研究者則有學術倫理與格式、程序上的要求。這些要求,從來都不是中性的,而是要你主動變成他們可以壓縮、可以理解、可以消費、可以取用、可以兌換另外一些東西的版本。

把這四件事疊起來,「低調」就不只是我一個人的個人偏好。
某種程度上,它是這個位置上、這份工作、這個十年,能夠存活下來的必要條件。

而即使越過這些結構條件,假設我願意更高調——「被看見」本身還有它自己的代價。

過去十年,我也有幸接過一些訪談邀約、合作邀請、媒體探詢。多數的共通點是,要求我進入一個既定的敘事框架。

露臉。這是基本要求。要露臉、要有人臉、要有可以做成reels、被轉發和截圖的視覺記憶點。

然後是內容上的要求:「平台上有什麼火辣的女女性愛故事?」「妳自己的個人故事、性偏好、性實踐才是我們想錄podcast的重點。」「女同志最常用的約炮話術?」「女同志都怎麼做愛?教大家正確知識嘛!」「所以跨性別有OO嗎?」

這些問題本身不全代表著惡意。它們是注意力經濟訓練出來的反射動作,在流量邏輯的訓練之下,大家都知道什麼樣的內容會被點開、被分享、被站隊。邀請合作方真心覺得這是在「幫我曝光」,他們真心誠意地認為「給你機會曝光」就是一種互惠。

但對我來說,這是對性少數獨立社群存在本身的完全錯解。
而且,一旦接受這種框架,就是選擇讓整個治理原則退讓。不被看見,作為一個人,我當然會覺得委屈。但如果要被看見的代價,是壓扁整個社群的複雜性,不尊重這個獨立社群的歷史與來處,也絲毫不考量這個社群面臨的風險與困境;只為了換取「被看見」——被誰看見?連對象是誰都無法掌握。看見後,然後呢?面對空泛的所謂曝光,萬一人潮湧入基礎設施、對社群造成的風險,我必須謹慎考慮。

所以我委婉地說「不。請你們調整一下機制,我們可以再談」。
然後注意力經濟的市場機制,會立刻判定我「沒有合作價值」「不夠有趣」「不會做行銷」「太難搞」。這裡面包含了某些主流的同志運動組織,因為他們也在苦苦掙扎著,需要「被大眾看見」。

因此,我完全接受一切難搞的評價。
在注意力經濟的邏輯裡,我確實是如此。這是我有意識的、自主的選擇。

沈默的雙面代價

另一方面,沈默也不是「自命清高的選擇」,反而是因為,已經說過了,結果更糟。
很多次、每一次。

困境會被武器化,社群內部長期有惡意使用者,反覆測試機制、進行破壞。只要堅持匿名、不收集個資、不公審的一天,這些惡意都不會消失。

很難理解的話,可以想像一下大家常批評的民主自由的弊病/代價,大概就能想像一二。

完全沉默,會被抹除。大肆發聲,會被武器化。
十年是在這之間,每天重新小心維持平衡。失衡一次,賠上的不只自己,很可能就是另一個社群成員的際遇;順利穩住,也只是換到下一個明天的喘息。沒有拍拍好棒,沒有安心解方,你辛苦了也無法當飯吃。

你不可能兩面討好,你終究要做選擇。並且承擔那個選擇。

光是寫這篇之前,我都花了數小時做必要的安全盤點。這不是一時衝動,也不是把社群或身邊的人推向不必要的風險。

你所謂的數位包容性:兩種「不被看見」的交會

這次事件最值得寫下的地方,並不是簡單的「臺灣人被排除」這個結論,而是它讓兩種「不被看見」在同一份名單上相撞、交會了。

一種是我個人這十年的不被看見——拒絕注意力經濟市場機制之後,所必須支付的代價。
另一種是臺灣作為一個主體的不被看見——國際組織裡的貶低改名、邦交斷裂、過境被拒、連公民社會的國際參與場合,都成為交易籌碼。

這兩種看起來很不一樣。一個是注意力經濟,和品牌行銷邏輯:另一個則是地緣政治,與外交資本的盤算。但兩者的底層邏輯,本質其實非常類似:決定誰可以出現、誰必須消失、出現的條件是什麼。

中華人民共和國對 RightsCon 主辦方提出的條件,包含現場參與和線上參與都不能讓「臺灣參與者、風險社群」出席。我不知道對方是否知道我這個位置的所有工作,也無從判斷我是不是這次行動的特定指向。

而這次事件把這兩個層次疊起來,疊在我這個位置上:性少數獨立社群站長 + 國際數位治理會議其中一個議程的臺灣主持者 + 高風險議題 + 平台治理研究——這四個東西交疊出來的人,不多。這讓我意識到:自己正在這幾個條件的多重交集上。

RightsCon 這場會議,本身的核心是探討全球公民社會在數位時代的包容性。當「繼續舉辦」的條件是「連『線上』,都不能讓臺灣人/風險社群/某些議題參加」——那麼,我們可以更清晰的看見數位包容性活生生、血淋淋的真實:「誰有資格」「用什麼條件」「可以怎麼使用、不能怎麼用」這些問題的複雜總和。

只是這次,這些問題,是由一通外交電話、政府閉門會議,甩出一個答案;留下其他上千名錯愕的國際數位人權倡議者、學者、研究者、政策制定者、社群實踐者在門外、國境之外。

最後

在收到消息後,立刻停下其他事情緊急寫這篇,是因為感受到一個少有的「開口」。

多年來我盡可能安靜,是因為法規、平台治理、議題特殊性、隱私等多方交織的策略選擇。但在這次事件,若連被排除的人都沈默不語,那他人強加的敘事就會成為歷史定稿。

書寫無法直接對抗排除,但書寫,可以逼迫排除無法假裝一切沒有發生。

「不被看見」可以是策略、是選擇、是拒絕為了被盲目、錯誤地看見而犧牲自己的複雜性。
「不被看見」也可以是被別人替你決定的命運。

過去十年,我跟濡沫的工作一直在前者之中。這次,我把後者與我的一部分故事也記下來。

在看見與不被看見之間,你也可以有你的選擇。