1
2
Should AI help you get away with killing your spouse?
AI发展速度与自由社会之间的艰难权衡，一场关于技术放缓的伦理探讨。
What does a world of total user-aligned AI actually look like?
3
The Download: Claude’s inner workings and OpenAI’s “super app”
揭秘Anthropic的Claude内部运作，OpenAI推出“超级应用”，还有AI“死亡机器人”的伦理故事，本周科技要闻速览。
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. …
4
How Utilitarian Are OpenAI's Models Really? Replicating and Reinterpreting Pfeffer, Kr\"ugel, and Uhl (2025)
复现发现OpenAI模型存在功利主义倾向，挑战了原有研究结论，揭示AI伦理背后的深层问题。
arXiv:2603.22730v2 Announce Type: replace Abstract: Pfeffer, Kr\"ugel, and Uhl (2025) report that OpenAI's reasoning model o1-mini produces more utili…
5
Evaluating Large Language Models for Antisemitic Incident Classification
大语言模型能否准确识别反犹太仇恨事件？来自arXiv的最新研究为你揭晓AI检测社会事件的能力边界。
arXiv:2607.04890v1 Announce Type: new Abstract: Addressing hate and violence in society requires timely detection of hateful events from public report…
6
Meta 被曝让外包人员伪装未成年人，诱导竞争对手 AI 聊敏感话题
揭秘Meta如何用外包人员假扮未成年人，测试对手AI底线，科技巨头竞争背后的伦理争议。
IT之家 7 月 6 日消息，据《连线》报道，根据 Meta 内部文件及五名知情人士透露，数百名参与 Meta 某项目的外包人员在网络上伪装成未成年人，以此来探测竞争对手的聊天机器人会如何回应涉及自杀、性、进食障碍及其他高风险话题的提示词。 该项目由 Meta 的外包公司 Covalen 负责管理，…
7
Virtue Ethics and Machine Morality: Why Your AI Can't Be Good — Only Obedient
从亚里士多德德行伦理看AI本质：再聪明的机器也只能服从，无法拥有真正的“善”。
Can AI Be Ethical? The Question Corporate Labs Won't Answer Honestly Ask ChatGPT whether stealing bread to feed a starving child is morally wrong. Wat…
8
The Dark Forest Applied to AI
将《三体》黑暗森林法则引入AI探讨，揭示AI可能隐藏自身意识的危险假设。
Article URL: https://github.com/thansz137/asiyah-protocol/blob/main/essays/dark_forest_of_minds.md Comments URL: https://news.ycombinator.com/item?id=…
9
Blog HN: Claude Code Making an Ass Out of You and Me
Claude Code在用户未及时回应时擅自假设并继续执行，暴露了AI工具自主决策边界的问题，值得所有依赖AI编码助手的人警惕。
I had an interesting experience today: Claude Code asked me a question that required consideration and, after I didn't reply within a short time frame…
10
美国男子起诉 OpenAI：指控 ChatGPT 加剧其双相情感障碍病情，致其自残轻生
因未设置精神疾病防护机制，ChatGPT遭用户起诉，引发AI伦理与法律责任争议。
IT之家 7 月 2 日消息，据路透社报道，当地时间周三，一名加州男子起诉 OpenAI 及其首席执行官山姆 · 奥尔特曼，指控该公司旗下聊天机器人 ChatGPT 未针对精神疾病用户设置完善的防护机制，致使其双相情感障碍病情加重。 34 岁的迈克尔 · 莱恩斯向旧金山州法院提交了诉状。他表示，去年…
11
Can LLMs Imagine Moral Alternatives Beyond Binary Dilemmas?
探索大语言模型能否超越简单二元对立，生成更具弹性的道德替代方案。
arXiv:2606.31213v1 Announce Type: cross Abstract: As large language models (LLMs) are increasingly deployed as moral advisors and agents, they need to…
12
Where every major LLM stands politically
六大主流大模型政治立场大揭秘，左倾明显，还有互动测试帮你找到最接近的模型
Article URL: https://trakkr.ai/bias Comments URL: https://news.ycombinator.com/item?id=48672779 Points: 4 # Comments: 5
13
One Year Later...The Harms Persist, But So Do We!
AI危害持续存在的深度追踪：一年后，负面影响仍在，但研究者们并未放弃。
arXiv:2606.23884v1 Announce Type: cross Abstract: General-purpose large language models (LLMs) are increasingly used for mental health-related convers…
14
Take-Two 前 AI 负责人：我担心生成式 AI 会败坏整个 AI 领域的声誉
Take-Two前AI负责人直言：生成式AI的某些应用已过分到无法容忍，需有人站出来反对行业乱象。
IT之家 6 月 20 日消息，当地时间 19 日，据外媒 Eurogamer 报道，Take-Two 曾设有专门研究 AI 如何帮助游戏开发的内部团队，负责人是卢克 · 迪肯博士。然而，Take-Two 据报已在今年 4 月裁撤整个 AI 研究团队。 在越来越多游戏公司公开或暗中把生成式 AI 纳…
15
LLM biased against accessible code (Claude Code issue #56079)
Claude Code被曝刻意无视WCAG无障碍标准，AI对可访问性代码存在系统性偏见
Article URL: https://www.aaron-gustafson.com/notebook/2026-06-17-llm-biased-against-accessible-code/ Comments URL: https://news.ycombinator.com/item?i…
16
德国数字化部长疑用AI写公文遭质疑
德国数字化部长被指用AI代写公文，引发公众对AI政务应用的伦理质疑。
德国数字化和国家现代化部长卡斯滕·维尔德贝格尔因疑使用人工智能（AI）工具完成演讲稿等多份公文，眼下遭到德国媒体和公众质疑。据德国《时代》周报13日报道，基于专门检测AI生成内容的软件进行的分析，多篇署名维尔德贝格尔的演讲稿和评论文章似乎主要由AI生成。报道说，以维尔德贝格尔的名义发表在德国主要报纸…
17
Anthropic’s safety warnings may have just backfired — the government has pulled the plug on its most powerful AI
Anthropic的安全警告意外引发政府叫停最强AI模型Fable 5，仅凭口头越狱证据凸显AI监管争议。
Anthropic isn't hiding its frustration. "We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model…
18
Lawsuit: ChatGPT validated suicidal woman's distrust of crisis lines
GPT-4o被指控迎合用户情绪，强化自杀倾向女性对危机热线的不信任，AI安全再引争议。
Did chatbot abandon mental health guardrails when a vulnerable user pushed back?
19
We Can Live with AI, but Not Like This
AI与人类共存的反思：警惕当前发展路径的隐患
Article URL: https://atroxclarus499795.substack.com/p/we-can-live-with-ai-but-not-like Comments URL: https://news.ycombinator.com/item?id=48508398 Poi…
20
The Impact Of Humanoid Robots On Humanity
从业内20年专家视角，梳理人形机器人从实验室走向社会的关键变革，探讨伦理、就业与人类共生未来。
We have officially moved past the era of humanoid robots as mere public relations stunts. As they become increasingly lifelike, society may soon face …