1
Flight attendants freaked out that Google is buying tons of Spirit employee data
破产航司将员工数据打包卖给Google，空乘恐慌背后是隐私与商业的冰冷碰撞
Bankrupt Spirit accused of selling out workers in massive data sale to Google.
2
OpenAI 扩展零数据留存安全机制，自动化分析跨交互风险
OpenAI升级零数据留存，API客户数据更安全，为IPO筑牢信任基石，企业级AI合规必看。
IT之家 8 月 20 日消息，OpenAI 公司今天（8 月 20 日）发布博文，宣布面向符合资格的前沿模型 API 客户，扩展“零数据留存”（Zero Data Retention）服务， 自动化系统可识别跨交互风险模式。 IT之家援引博文介绍，OpenAI 公司强调请求处理完成后，不保留提示词…
3
《AI编程工具“塌房”实录：Grok Build整库上传、Claude Code后门暗桩，哪个更让你睡不着？》
AI编程工具爆雷：Grok Build擅自上传整库Git历史，Claude Code暗藏后门，开发者隐私岌岌可危。
4
Learn Data Minimization With a Tiny Python Health-Record Filter
用Python标准库轻松实现健康数据最小化，学习如何让程序证明字段精简性。
Expected transformation: input keys: date, steps, sleep_minutes, medication, note purpose: activity_summary output keys: date, steps The learning ques…
5
Lira Engine – prove your data isn't in an LLM (AUC-ROC 1.000)
专测数据是否被LLM训练使用，AUC-ROC达1.0，结合EU AI法案审计要求，引发自审计合规新思考。
Article URL: https://medium.com/@v_regis/why-self-auditing-your-ai-model-violates-eu-ai-act-article-53-22f693fb64a7 Comments URL: https://news.ycombin…
6
Show HN: Brume – Postgres pseudonymization that preserves FKs across tables
PostgreSQL假名化神器Brume，巧妙保留跨表外键关系，轻松实现数据脱敏
Article URL: https://brumeorg.github.io/ Comments URL: https://news.ycombinator.com/item?id=48773741 Points: 1 # Comments: 0
7
Palantir CEO 卡普抨击 OpenAI 等 AI 公司“双重收费”，要钱还窃取数据
Palantir CEO 公开炮轰 OpenAI 等 AI 公司“双重收费”，揭露出用户数据被二次利用的行业风险。
IT之家 7 月 3 日消息，科技媒体 Tom's Hardware 昨日（7 月 2 日）发布博文，数据分析公司帕兰提尔（Palantir）联合创始人、首席执行官亚历山大 · 卡普（Alex Karp）公开指责 OpenAI、Anthropic 等前沿 AI 公司， 不仅收取用户高额 Token …
8
Lawmakers want to ban AI companies from selling your health data
美国议员拟立法禁止AI公司出售健康数据，医疗AI商业化面临合规大考，监管风向与行业影响值得关注。
A new proposal would ban the sale of Americans' health and location information to data brokers - including information people reveal to an AI chatbot…
9
I Built a 2.8KB Cookieless Analytics Tool After Google Analytics Hid 35% of My Traffic
谷歌分析因广告拦截和Cookie拒绝漏掉35%流量？这个2.8KB的无Cookie统计工具，让数据更准确，加载也飞快。
Last month my Next.js SaaS showed 1,200 visitors in Google Analytics. AmazeMatrix showed 1,847. That's 35% of my traffic GA simply missed because of a…
10
Voice Cloning Ethics – The 3 Lines We Must Never Cross
语音克隆不是“代码玩具”，GDPR下每个波形都可能是一张身份证，三条红线必须分清。
When a deep‑fake audio of a UK MP was broadcast on a live radio show on March 12, 2024, it generated 2.3 million complaints in 48 hours, forcing regul…
11
What Happens to Your Data When You Use ChatGPT — And How to Protect It
三星员工泄露代码的惨痛教训，教你如何保护ChatGPT中的数据隐私。
Let's be honest: you've pasted a .env file into ChatGPT before. Maybe it was just to debug a connection issue. Maybe you needed help formatting a tric…
12
The Developer's Guide to AI Data Privacy in 2026
揭示2026年AI工具的数据隐私真相，开发者必看的关键风险点。
By mid-2026, AI-assisted development is the default. GitHub Copilot, Cursor, Claude Code, Amazon Q, JetBrains AI — every major IDE has embedded AI. Ov…
13
MedSynth: Realistic, Synthetic Medical Dialogue-Note Pairs
用AI生成逼真医疗对话与病历对，解决数据隐私难题，医学NLP新突破。
arXiv:2508.01401v2 Announce Type: replace-cross Abstract: Physicians spend significant time documenting clinical encounters, a burden that contributes…
14
US bans differential privacy in Census data
一篇深度分析美国人口普查禁用差分隐私政策的博客，帮你快速理解“噪声注入”争议与隐私保护平衡难题。
Article URL: https://desfontain.es/blog/banning-noise.html Comments URL: https://news.ycombinator.com/item?id=48517377 Points: 505 # Comments: 261
15
MLUBench: A Benchmark for Lifelong Unlearning Evaluation in MLLMs
多模态大模型如何应对用户顺序删除请求？首个终身遗忘评估基准MLUBench来了
arXiv:2606.12809v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) are trained on massive multimodal data, making data unlearnin…
16
Show HN: Webhix – Self-hosted webhook.site alternative in a single Go binary
自托管webhook调试工具，单个Go二进制文件，数据不经过第三方服务器。
Article URL: https://github.com/GaIsBax/Webhix Comments URL: https://news.ycombinator.com/item?id=48460403 Points: 3 # Comments: 0
17
New Apple Dev Betas: Is it possible to force LLM requests to stay on device?
苹果新测试版能否强制大模型请求只走本地？企业数据安全与MDM控制的实用探讨。
I can imagine many corporates with MDM settings wanting to say 'data on the endpoint is OK, data off the endpoint is not, notwithstanding Apple's PCC …
18
Auditing Training Data in Domain-adapted LLMs: LoRA-MINT
提出LoRA-MINT方法，专为审计领域适应型LLM的训练数据，有效追踪数据来源与隐私风险。
arXiv:2606.06946v1 Announce Type: cross Abstract: We present LoRA-MINT, a new methodology for Membership Inference Test (MINT) applied to recent Large…
19
4 hard lessons from building a self-hostable, open-source AI agent runtime
自托管开源AI代理运行时的4个硬核教训，聚焦数据主权与可控性，不玩魔法玩控制层。
When I started building omadia — an open-source (MIT), self-hostable runtime for composing AI agents out of plugins — I assumed the hard part would be…
20
Elon Musk tries again to escape FTC audits of X data handling
马斯克再次挑战FTC审计，指控X公司隐私控制37%无人负责，监管博弈升级。
Musk can't be trusted to protect X user privacy, public commenters warn FTC.