1
How to Steal an AI Model’s Private Thoughts
破解加密推理闭环，看研究团队如何窥探AI私有思考，安全边界再受拷问
In August 2026, a team at MATS Research, the ELLIS Institute Tübingen, and the Max Planck Institute for Intelligent Systems wanted to test whethe…
2
Claude Cowork finally remembers what you told the app in chat
Claude Cowork终于能记住聊天中的信息了，还默认保护敏感数据，可手动开启个性化记忆。
Anthropic is giving Claude a shared memory across chat and Cowork, so users no longer have to repeatedly brief the AI on projects, preferences, and ot…
3
Nitter project received cease and desist
Nitter开源项目收到停止函，公共实例全军覆没，隐私访问Twitter的替代方案再遭重击。
Article URL: https://github.com/zedeus/nitter/issues/1442 Comments URL: https://news.ycombinator.com/item?id=49437283 Points: 255 # Comments: 173
4
Apple rescues Hide My Email feature from the privacy scrap heap
苹果官方确认“隐藏邮件地址”功能保留，隐私党安心了，隐私保护不掉队。
Apple says it will no longer ditch using its icloud.com domain for hiding people's email addresses.
5
X 发律师函要求关停 Nitter，免登录浏览 X 帖文项目被迫下线
无需登录即可阅读 X 帖文的开源项目，因律师函被迫下架，隐私浏览体验成绝响
IT之家 8 月 26 日消息，Nitter 是一个开源项目，允许用户无需登录，甚至无需打开 X 应用，就能阅读 X 上的帖子。如今，X 已向 Nitter 发出停止侵权并终止相关活动的律师函，要求其关闭。该消息通过 Nitter 网站上的一则简短公告对外公布，此前 X 已经尝试通过技术手段让 Ni…
6
Abby Steele: An Offline AI Personality Powered by a Local LLM and Chess Engine
想体验不联网的AI棋友？Abby Steele用本地大模型驱动，下棋聊天两不误，隐私不越界。
Article URL: https://www.abbysteele.com Comments URL: https://news.ycombinator.com/item?id=49430970 Points: 1 # Comments: 0
7
WhatsApp tightens account security with stronger two-step verification and more
WhatsApp升级两步验证，账户安全再加固，保护你的聊天隐私不被破解。
WhatsApp’s two-step verification previously relied on a six-digit PIN, but now users can choose a longer, alphanumeric password with special character…
8
Stress Testing Unlearning Algorithms
系统检验机器遗忘算法在极端压力下的鲁棒性，为隐私保护研究划出新基准。
arXiv:2608.22527v1 Announce Type: new Abstract: Recently, machine unlearning, the removal of specific training data influence from a model, has gained…
9
SplitLite: Low-Rank Residual Compression for Split Learning
用低秩残差压缩给分割学习“减负”，通信开销大减，隐私与效率兼得的新思路。
arXiv:2608.23018v1 Announce Type: cross Abstract: Federated fine-tuning of on-device large language models (LLMs) faces a significant computing burden…
10
I Gave ChatGPT Access to My iMessage Conversations, and It Was More Useful Than I Expected
把iMessage交给ChatGPT，意外收获：聊天记录变私人助理，Mac用户即刻可试。
Getting ChatGPT connected to your Apple Messages comes with privacy trade-offs—but it has its uses.
11
Instinct’s powerful AI assistant is raising privacy and security concerns
AI助手如魔法般强大，但隐私安全漏洞令人心惊，值得每个用户警惕。
Early testers are raving about what Instinct can do, but some say the AI assistant’s sweeping access, broad terms and ability to act on users’ behalf …
12
Bring Your Own Key: Building an AI Portal That Never Touches the Billing
自带API密钥的AI聚合门户，让你绕过订阅加价直接调用各大模型，隐私与成本兼得。
Bring-your-own-key, not another subscription Every few months another "all AI models in one place" product shows up, and almost all of them work the s…
13
Daimon – Local Privacy LLM
本地运行隐私优先的LLM服务，敏感文本不触达外部模型，开源可自部署。
Article URL: https://github.com/ar0per0/Daimon Comments URL: https://news.ycombinator.com/item?id=49408697 Points: 1 # Comments: 0
14
OpenCyvis: An open-source AI phone agent that runs your own LLM
开源AI手机代理，运行自有LLM，手机操作全透明，告别商业黑盒。
Article URL: https://github.com/opencyvis/opencyvis-phone Comments URL: https://news.ycombinator.com/item?id=49381419 Points: 1 # Comments: 0
15
Show HN: Encrypt messages locally into self-contained links
本地加密生成自毁链接，信息只存于链接中，无需服务器留存，阅后即焚保护隐私。
All encrypting and signing uses post-quantum primitives. This is a static page; nothing is ever sent to the server. Comments URL: https://news.ycombin…
16
Building an AI Translator Keyboard for iOS and Android: Privacy, Latency, and UX Lessons
从隐私、延迟到UX，AI翻译键盘开发的真实教训，避免常见设计误区。
Translating a message on a phone often means leaving the conversation, opening a translator, pasting the text, copying the result, and returning to th…
17
As demand for Meta AI glasses explodes, it’s harder to avoid creepy recordings
AI眼镜热卖却难防偷拍，免费检测应用能否终结隐私恐慌？
Ars looks at Zuckoff, the latest free app detecting Meta AI glasses amid privacy backlash.
18
Offering Zero Data Retention for frontier models
OpenAI为前沿模型API提供零数据保留，并预览隐私安全处理，兼顾AI安全与数据隐私。
OpenAI reaffirms Zero Data Retention for eligible API customers and previews Private Safety Processing for advanced AI safety without compromising dat…
19
Offering Zero Data Retention for frontier models
OpenAI强化API隐私承诺，零数据保留加私密安全处理，企业用AI更安心
OpenAI reaffirms Zero Data Retention for eligible API customers and previews Private Safety Processing for advanced AI safety without compromising dat…
20
Redakto - The Incognito Tab for LLMs
大模型也需“无痕模式”，Redakto为LLM对话提供隐私保护新方案。
arXiv:2608.18260v1 Announce Type: new Abstract: Large Language Models (LLMs) are being increasingly used in everyday applications. A major challenge i…