Author: RadarAI Editorial
Editor: RadarAI Editorial
Last updated: 2026-11-11
Review status: Editorial review pending
Brief
速报
官方
AI动态
开源
This week the AI industry saw a dual rhythm of dense model releases and accelerated regulatory implementation: Google DeepMind launched Gemini 4 Carbon, its strongest public model to date, followed closely by Alibaba's Qwen4-Max and Baidu's ERNIE 5.5 Flash Preview, with leading vendors continuing to push coding, multi-agent, and long-context capabilities [0][3][10]; meanwhile, New York's RAISE Act regulatory platform opened for registration, bringing OpenAI, Anthropic ...
Editorial standards and source policy: Editorial standards, Team. Content links to primary sources; see Methodology.
## 🔍 Core Insights
This week the AI industry saw a **dense period of model releases** and **accelerated regulatory implementation**: **Google DeepMind** launched **Gemini 4 Carbon**, its strongest public model to date, with **Alibaba Qwen4-Max** and **Baidu ERNIE 5.5 Flash Preview** following closely, as leading vendors continued to push coding, multi-agent, and long-context capabilities [0][3][10]; meanwhile, **New York's RAISE Act** regulatory platform opened for registration, bringing large developers such as **OpenAI and Anthropic** into a 72-hour security incident reporting framework, substantially increasing compliance pressure [2]. On the capital side, the **TSMC** board approved an investment of up to **US$1 billion** in Anthropic and OpenAI, further deepening the ties between the industrial chain and the model layer [13].
## 🚀 Key Updates
- **Gemini 4 Carbon officially launched** [0]: Google DeepMind's strongest public model, focused on complex coding, cybersecurity, and multi-agent tasks; regular users need a Google AI Ultra subscription.
- **New York opens RAISE Act regulatory platform** [2]: Frontier AI developers with annual revenue over US$500 million must complete registration by 2027 and report security incidents within 72 hours.
- **Alibaba Qwen4-Max Preview launched early** [3]: A preview is open before the official release, with a limited-time 10% discount on token pricing; it will be built into Qoder, Qwen Office, and QwenStudio.
- **Baidu ERNIE 5.5 Flash Preview released** [10]: 280B total parameters, 14B activated parameters, supports 1M context, with open-source weights before November 25.
- **Tencent Yuanbao Search officially launched** [8]: Can search WeChat Official Accounts and Channels, connected to 157 authoritative sources, with ads clearly distinguished from answers.
- **TSMC approves up to US$1 billion investment in Anthropic and OpenAI** [13]: Purchasing shares in both companies at IPO issue price; October revenue grew 60% year over year over the same period.
- **Anthropic's new policy penetrates and bans Chinese teams** [20]: The policy for supported regions takes effect, dealing a severe blow to Chinese AI teams expanding overseas and overseas entities with Chinese backgrounds.
- **UK OfS warns AI threatens the credibility of degrees** [4]: A survey says teachers can barely judge whether assignments reflect students' true level, calling for a return to traditional exams and in-person presentations.
## 🔗 Sources
[0] Starting today, Gemini 4 Carbon is officially launched, but not everyone can use it — https://www.bestblogs.dev/article/0c99d7db6c?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item
[2] New York opens the RAISE Act regulatory registration platform; OpenAI, Anthropic, and others must report security incidents within 72 hours — https://aihot.news/items/kgo13ml86e77sjcbz0si6rzop
[3] Alibaba Qwen4-Max preview launch and limited-time offer — https://aihot.news/items/bqxrrd9esj59au6k554hho17x
[4] UK Office for Students survey says AI makes it nearly impossible for teachers to judge whether coursework reflects students' true level — https://aihot.news/items/q3ej22t537eddl1zldw21q3g1
[8] Tencent Yuanbao Search officially launched: can search WeChat Official Accounts and Channels, with ads clearly distinguished from answers — https://aihot.news/items/j3g8gwmdk3hhhoz8ifkfdseoz
[10] Baidu Wenxin releases ERNIE 5.5 Flash
This week the AI industry saw a dense period of model releases and accelerated regulatory implementation: Google DeepMind launched Gemini 4 Carbon, its strongest public model to date, with Alibaba Qwen4-Max and Baidu ERNIE 5.5 Flash Preview following closely, as leading vendors continued to push coding, multi-agent, and long-context capabilities [0][3][10]; meanwhile, New York's RAISE Act regulatory platform opened for registration, bringing large developers such as OpenAI and Anthropic into a 72-hour security incident reporting framework, substantially increasing compliance pressure [2]. On the capital side, the TSMC board approved an investment of up to US$1 billion in Anthropic and OpenAI, further deepening the ties between the industrial chain and the model layer [13].
🚀 Key Updates
- Gemini 4 Carbon officially launched [0]: Google DeepMind's strongest public model, focused on complex coding, cybersecurity, and multi-agent tasks; regular users need a Google AI Ultra subscription.
- New York opens RAISE Act regulatory platform [2]: Frontier AI developers with annual revenue over US$500 million must complete registration by 2027 and report security incidents within 72 hours.
- Alibaba Qwen4-Max Preview launched early [3]: A preview is open before the official release, with a limited-time 10% discount on token pricing; it will be built into Qoder, Qwen Office, and QwenStudio.
- Baidu ERNIE 5.5 Flash Preview released [10]: 280B total parameters, 14B activated parameters, supports 1M context, with open-source weights before November 25.
- Tencent Yuanbao Search officially launched [8]: Can search WeChat Official Accounts and Channels, connected to 157 authoritative sources, with ads clearly distinguished from answers.
- TSMC approves up to US$1 billion investment in Anthropic and OpenAI [13]: Purchasing shares in both companies at IPO issue price; October revenue grew 60% year over year over the same period.
- Anthropic's new policy penetrates and bans Chinese teams [20]: The policy for supported regions takes effect, dealing a severe blow to Chinese AI teams expanding overseas and overseas entities with Chinese backgrounds.
- UK OfS warns AI threatens the credibility of degrees [4]: A survey says teachers can barely judge whether assignments reflect students' true level, calling for a return to traditional exams and in-person presentations.
🔗 Sources
[0] Starting today, Gemini 4 Carbon is officially launched, but not everyone can use it — https://www.bestblogs.dev/article/0c99d7db6c?utm_source=rss&utm_medium=feed&utm_campaign=resources&entry=rss_article_item
[2] New York opens the RAISE Act regulatory registration platform; OpenAI, Anthropic, and others must report security incidents within 72 hours — https://aihot.news/items/kgo13ml86e77sjcbz0si6rzop
[3] Alibaba Qwen4-Max preview launch and limited-time offer — https://aihot.news/items/bqxrrd9esj59au6k554hho17x
[4] UK Office for Students survey says AI makes it nearly impossible for teachers to judge whether coursework reflects students' true level — https://aihot.news/items/q3ej22t537eddl1zldw21q3g1
[8] Tencent Yuanbao Search officially launched: can search WeChat Official Accounts and Channels, with ads clearly distinguished from answers — https://aihot.news/items/j3g8gwmdk3hhhoz8ifkfdseoz
[10] Baidu Wenxin releases ERNIE 5.5 Flash
← Back to Updates