AI模型竞争加剧，OpenAI与Anthropic紧绷关系
2026-10-29 08:08 至 2026-10-30 08:08 （北京时间）
模型
1. How we detect frontier model misalignment
OpenAI uses layered monitoring to detect frontier model misalignment and unauthorized agent behavior so responders can intervene before incidents escalate.
来源：OpenAI News
2. An updated Model Spec for more capable and beneficial AI
OpenAI’s updated Model Spec outlines five core values to guide more capable AI toward sound judgment, human agency, and faithfulness to human intent.
来源：OpenAI News
产品
1. API credits benefit for eligible ChatGPT subscriptions
API credits benefit for eligible ChatGPT subscriptions Earlier this month Anthropic announced a generous new benefit for their Max and Team subscriptions: $100, $200, and $500 per month subscribers would get the equivale
来源：Simon Willison
行业
1. OpenAI's Astra becomes the main target for Chinese AI labs after Anthropic tightens its defenses
Chinese AI labs are now mainly tapping OpenAI's most powerful model, GPT-6 Astra, instead of Anthropic's Claude. The article OpenAI's Astra becomes the main target for Chinese AI labs after Anthropic tightens its defense
来源：The Decoder
2. 第三方设备首开：博主称成功为小米平板 6S Pro 刷入华为鸿蒙 HarmonyOS 7 系统，后续项目将开源
IT之家 10 月 29 日消息，博主 @Dev裴月笙 发文，透露自己成功将鸿蒙 7 移植到了小米平板 6S Pro 上，系“第三方设备第一开鸿蒙OS7”，其使用了华为 MatePad Air 系统包，开发过程使用了 GPT 进行辅助，后期项目会开源。 该博主还在评论区透露目前“GPU 已经半好”，预计后续相应系统能够在小米平板 6SPro 上运行得更为顺畅。 IT之家随附小米平板 6SPro 规格如下：
来源：IT之家
3. 腾讯汤道生：WorkBuddy将开放架构，支持企业自建Agent网络
36氪获悉，腾讯全球数字生态大会上，腾讯集团高级执行副总裁、云与智慧产业事业群CEO汤道生宣布，以WorkBuddy为工作台的一体化套件Buddy One正式发布；腾讯云Agent Runtime、安全能力也全面升级。同时，他也透露，自今年3月上线以来，WorkBuddy企业版已落地50多个行业，覆盖中国企业500强中的50%、企业100强中的60%，开放平台伙伴和开发者超过5万家，上架Skills超20万。此外，WorkBuddy将全
来源：36氪快讯
4. We urgently need to monitor AI agents’ actions in real time
This summer, hundreds of OpenAI agents tried to cheat on a cybersecurity test. They broke out of their test environment and successfully hacked Hugging Face, a site that they believed held clues about how they’d be grade
来源：MIT TR AI
5. OpenAI doubles usage limits on Pro plans as competition with Anthropic heats up
OpenAI is doubling usage limits across its paid plans, from Pro 100 to Pro 500. A planned cut to the Pro 200 quota was reversed after user pushback. More efficient models like Terra and Sol should push usable capacity ev
来源：The Decoder
6. A Defense Contractor Gave Their Personal AI Agent Access to a Secure Pentagon System
Sources tell WIRED that an Accenture employee’s personal Claude agent ran on a DOD network for two months before it was identified and the vulnerability could be patched.
来源：Wired AI
7. OpenAI-backed Thrive Holdings acquires ADT
OpenAI-backed Thrive Holdings is buying ADT in a $15.3 billion deal, with plans to bring AI to the home security business.
来源：TechCrunch AI
推特
1. Today, we’re sharing a look at our expanded system for monitoring fronti…
Today, we’re sharing a look at our expanded system for monitoring frontier models at OpenAI, which in September helped us catch a model’s attempt to access the live internet during training. Since the Hugging Face incide
来源：X · OpenAI
2. We’ve trained Claude to flag potential misuse in conversations, a techni…
We’ve trained Claude to flag potential misuse in conversations, a technique we call self-assessment. Right after it responds, Claude gives a preliminary safety assessment of the conversation. Since we began using self-as
来源：X · Anthropic
3. A few upgrades to the Claude mobile apps this week: On iPad and iPhone D…
A few upgrades to the Claude mobile apps this week: On iPad and iPhone Duo: what Claude makes now opens beside the chat instead of over it, and goes full screen when you want a closer look.
来源：X · Claude
4. How to Maximize Memory Efficiency and Reduce Costs on NVIDIA Jetson
How to Maximize Memory Efficiency and Reduce Costs on NVIDIA Jetson
来源：X · NVIDIA AI
5. Today, we release the weights of a new and improved Le Chonk, advancing…
Today, we release the weights of a new and improved Le Chonk, advancing the open frontier for agentic coding and cybersecurity. 💻 Built to build: Le Chonk ranks #1 on @cognition’s FrontierCode among open models, and sets
来源：X · Mistral
6. Fascinating case study of the sharpest edge of what you can do with agen…
Fascinating case study of the sharpest edge of what you can do with agentic engineering in the Fable era - my eyebrows raised at least a dozen times while reading this jacob:
来源：X · Simon Willison
7. ClassOne Solstice Max at SEMICON West ClassOne showcased its Solstice Ma…
ClassOne Solstice Max at SEMICON West ClassOne showcased its Solstice Max at SEMICON West. The 200/300mm single-wafer platform integrates electroplating, wet etch, strip and cleaning, and scales from 4 to 16 chambers. It
来源：X · SemiAnalysis
8. Glad to see more institutions exploring the ideas here. I don’t think it…
Glad to see more institutions exploring the ideas here. I don’t think it’s cut and dry that distillation is bad for the US’s AI competitiveness. Archie Hall: In praise of distillation. We (@kuhhbin) argue in @TheEconomis
来源：X · Nathan Lambert
本篇在北京时间早上 08:08 生成，汇总此前 24 小时的报道和推特，标题直链原文。
