Taipei | Claude and AI Ethics

一起來 Claude 社群活動參與討論吧！本活動由在地愛好者為所有熱愛用 Claude Code 打造產品的人所舉辦！

這次我們要做一些不一樣的事。我們將聚集一群對 AI 倫理與科技未來有興趣的人，一起深入討論一個嚴肅但相當重要的問題：

在 AI 技術滲入戰爭決策與地緣政治衝突的現在，負責任地開發 AI 究竟意味著什麼？這個決定權又該由誰來掌握？這次聚會將從兩個角度認真探討這個問題。

第一個角度是 Anthropic 目前與聯邦政府正在進行的訴訟，Anthropic 曾經是美國戰爭部最為倚重的人工智慧系統供應商，然而在近期的地緣政治衝突中，五角大廈要求 Anthropic 移除 Claude 在自主致命武器與大規模國內監控方面的使用限制，但 Anthropic 拒絕。之後，五角大廈將其列為「國家安全供應鏈風險」。相關法院文件，包括 Anthropic 提告的起訴書、Microsoft 以及 OpenAI、Google 工程師個人提交的法庭之友意見書——揭示了 AI 安全原則與國家權力之間的正面衝突。第二個角度是 Anthropic 針對 159 個國家共 81,000 人所進行的研究。該研究由曾參與集體智慧計劃（CIP）的研究員主導，目標是了解大家對於人工智慧的看法。該研究聚焦於 Claude 使用者對於人工智慧的看法，在龐大的研究規模之外，也引發研究方法與代表性的質疑。

這兩件事剛好都與 Anthropic 相關。將它們放在一起觀察，會發現一個更有意思的問題：如果Anthropic 與美國聯邦政府的訴訟，關乎的是「政府與企業誰有權為人工智慧設定界線」，那麼這項涵蓋 81,000 人的研究則回到更根本的議題，也就是人工智慧系統的界線不僅僅是政府與企業的爭端，而會對有著真實且彼此衝突感受的人產生實際的影響。

當一個由私人公司開發、並被數億人使用的模型，逐漸使某些關於「安全」或「倫理」的思考方式變得理所當然時，我們應該要求什麼樣的問責機制？

目前在討論人工智慧政策的制度與每天實際與人工智慧共處的人，往往並不在同一場對話之中。這也是我們希望透過這場活動來建構這個對話的原因。

由於本次聚會的主題性質，我們刻意縮減規模，以便讓參與者與講者能有更深入的討論。如果你對這個議題充滿熱情，請在報名時告訴我們你的想法。我們歡迎各種背景與不同觀點的參與者。

在聚會前，請閱讀以下相關資料，讓自己能更充分地參與討論：

Anthropic 對美國國防部的聯邦訴訟（連結）Microsoft （連結）及 OpenAI、Google 工程師個人提交的支持意見書（連結）Anthropic 的 81,000 人研究（連結）補充閱讀——端傳媒文章：Anthropic 與國防部之爭：誰有權決定AI的邊界？（連結）

講者介紹：

主持兼導讀 Peter Cui ：台大法律研究所，法律研究者，vTaiwan 社群貢獻者，關注人工智慧治理與規範，入選 ISF Global Fellow 2025。 國防產業與談 Anshuman Prasad：曾在 Palantir Technologies 任職 8 年，期間參與多項專案，包括為 Project Maven 開發邊緣人工智慧能力，以及疫苗配送與航太相關應用。目前為 Valinor 的工程團隊負責人。 科技治理與談 Nicole Chan：台灣國家通訊傳播委員會前主任委員，現為產業顧問與執業律師、ICANN ASO AC、人工智慧科技基金會董事長、台灣數位信任協會理事長。

其他細節：

午餐：由於我們將有一位海外講者視訊連線，聚會將於早上 9 點開始。討論將延續至午餐時間，我們會為每位參與者訂購便當，費用由我們負擔！請在報名時告知我們是否需要為你訂餐。語言：活動將交替使用英文與中文，請用你最自在的語言參與討論。

感謝所有讓這次活動得以實現的夥伴：

場地：g0v 社群空間活動籌備：Web3 For All 社群支持：Claude 社群團隊

如果你有興趣但無法出席，我們將在活動進行時開放直播，並在活動結束後與更廣泛的社群分享重點摘要。請注意，若你的報名已獲接受，但之後無法出席，請提前告知我們，以便將名額開放給其他人。未事先告知的缺席者將在未來活動中被列為較低優先順序。

本次聚會中所有表達的觀點均屬個人立場，不代表 Anthropic 之意見或立場。

–

Join a Claude Community Event, organized by local enthusiasts for everyone who loves building with Claude Code!

This time, we’re doing something a little different. We’re gathering a group of interested participants who want to discuss a serious but important topic: ethics surrounding AI and the future of technology.

When AI has infiltrated into war and geopolitics conflict, what does it actually mean to build AI responsibly and who gets to decide? This meetup takes that question seriously through two lenses.

The first is Anthropic's ongoing federal lawsuit against the U.S. Department of War. Anthropic is once the main AI model provider for Department of War. However, during the recent geopolitical crisis, when the Pentagon demanded Anthropic remove its restrictions on using Claude for autonomous lethal warfare and mass domestic surveillance, then designated Anthropic a "supply chain risk to national security" after it refused. The court filings — including supporting briefs from Microsoft, and from individual engineers at OpenAI and Google — reveal how AI safety principles collide with state power.The second lens is Anthropic's study of 81,000 people across 159 countries on what they actually want from AI. The research project is led by a researcher participating in Collective Intelligence Project (CIP). The research focus on understanding the viewpoints from AI users about AI governance. However, the scale and representativeness also raise concerns.

Both readings happen to come from Anthropic. Put side by side, they raise something more interesting: if the lawsuit is about who gets to set limits on AI, the 81,000-person study is a reminder that those limits affect real people with real and conflicting feelings about it.

When a model used by hundreds of millions of people normalizes certain ways of thinking about what's safe or ethical, what kind of accountability does that demand? The institutions debating AI policy and the people living with it daily are often not in the same conversation, and maybe they should be.

Due to the theme of this meetup, we are keeping the group smaller intentionally so we can have deeper discussions. If you’re passionate about this topic, please let us know why when you RSVP for the event. We welcome all backgrounds and diverse thoughts.

Before the meetup, please take a look at the relevant docs to participate more fully in the discussion:

Anthropic’s ongoing federal lawsuit against the U.S. Department of War (link)Supporting briefs from Microsoft (link) and individual engineers at OpenAI and Google (link)Anthropic’s study of 81,000 people (link)Supplementary reading of Initium Media article: Anthropic 與國防部之爭：誰有權決定AI的邊界？(link)

We also invite speakers with diversified backgrounds but with common interests and proficiency on AI governance.

Speakers:

Moderator & Discussant — Peter Cui: Graduate student at National Taiwan University College of Law, legal researcher, and contributor to the vTaiwan community. Focuses on AI governance and regulation. ISF Global Fellow 2025. Defense Industry Panelist — Anshuman Prasad: Formerly spent 8 years at Palantir Technologies, working on multiple projects including developing edge AI capabilities for Project Maven, as well as applications in vaccine distribution and aerospace. Currently leads the engineering team at Valinor. Tech and Governance Panelist — Nicole Chan: Former Chairperson of Taiwan’s National Communications Commission. Currently an industry advisor and practicing attorney, ICANN ASO AC, Chairperson of the Artificial Intelligence Foundation and Digital Trust Association in Taiwan.

Other details:

Lunch: We're joining a speaker dialing in from overseas so we will start our meetup at 9 AM. Since the discussions will go through lunch, we will order bento meals for every participant. Please let us know if you’d like us to order you a lunch when you RSVP for the event. It’s on us!Language: English and Mandarin will be used interchangeably. Use the language you feel most comfortable speaking in.

Thanks to all those who made this possible:

Venue: g0v community spacePlanning: Web3 For All communitySupport: Claude community team

If you’re interested but can’t make it, we will also be livestreaming during the event, and we'll share key takeaways with the wider community after the event. Please note that if we accept your spot and you can no longer attend, please let us know in advance so we can offer your spot to someone else. No-shows who don't communicate with us will be deprioritized for future events.

All views expressed at this meetup are those of the individual speakers and participants. We do not claim to represent Anthropic in any way.

–