2026-09-14
AI Briefings Daily Report.

China criticizes US AI leaders' safety warnings, calling them intimidation to lock in US advantage ↗

China has flatly rejected warnings about AI risks from Anthropic CEO Amodei and other U.S. AI leaders. Beijing's Foreign Ministry calls it "fearmongering," while the state-run Global Times accuses Amodei of waging a "silent AI Cold War." China's security minister isn't calling for a slowdown either but for faster AI infrastructure buildout. Trump also opposes any slowdown.

Report: Hillhouse Ventures partner Yan Wentao expected to become DeepSeek CFO ↗

DeepSeek's CFO (Chief Financial Officer) is about to take office, the person closest to the final candidate is Yan Wentao, partner at Hillhouse Ventures, who is currently in the process of resigning. (Phoenix Network Tech)

Tencent Yuanbao on AI homework tutoring: how parents can use AI detailed explanation to study with children ↗

The Cybersecurity Standardization Technical Committee released the 'Artificial Intelligence Security Governance Framework 3.0' ↗

At the opening ceremony of the 2026 National Cybersecurity Awareness Week, the National Information Security Standardization Technical Committee released the 'Artificial Intelligence Security Governance Framework 3.0'. The framework updates the risk classification, optimizes governance measures, and aims to enhance AI security governance capabilities and ensure AI benefits humanity.

Anthropic reports Houthi forces used Claude Code to develop missile guidance software ↗

Anthropic says operators in northern Yemen ran parallel Claude Code instances to design guidance software, simulate trajectories, and analyze a failed rocket t…

Sayash Kapoor uses the AI as Normal Technology framework to interpret the runaway incident: the OpenAI - Hugging Face incident was a control failure rather than a pure alignment crisis ↗

A middle ground between the cybersecurity and AI safety communities

OpenAI co-founder Brockman says AI development slowdown should be limited to frontier large models only ↗

OpenAI President Brockman said measures to slow AI research and development should apply only to frontier large models and should not restrict open-source and personal projects. He mentioned that after the Hugging Face incident, OpenAI has improved its safety mechanisms and called for international coordination.

CEOs of multiple US AI companies call for slowing R&D; AI concept stocks fall across the board ↗

CEOs of AI giants including Anthropic and OpenAI warn about AI risks and call for slowing development, triggering volatility in global stock market AI sectors. Tech stocks such as SoftBank and Kioxia plummet, while China and the US plan to hold AI safety dialogues.

Sam Altman calls for slowing AI development pace, but promises rapid progress will continue ↗

Sam Altman is doubling down on slowing AI development. OpenAI now runs safety checks before major training runs, and according to The Information, the company has been talking with Anthropic and Google for months about joint self-regulation.

Foreign Ministry responds to US AI company threat theory: spreading threat narratives only disrupts the global AI governance process ↗

In response to US AI companies claiming that China's AI development poses security risks, Foreign Ministry spokesperson Guo Jiakun stated that all parties should jointly promote open, inclusive, and beneficial AI development, warning that spreading threat narratives will only disrupt global governance processes.

Weibang Zhendian completes nearly 100 million yuan Series B financing, led by Daotong Investment, advancing commercialization of industrial inspection, clinical, and AI for Science ↗

Zhejiang Weibang Zhen Dian Optoelectronics Technology Co., Ltd. (hereinafter referred to as Weibang Zhen Dian) has completed a Series B funding round of nearly 100 million yuan. The round was led by Daotong Investment, with IDG Capital participating. The funds will primarily be used to commercialize products in industrial inspection, clinical applications, and AI for Science. Previously, Weibang Zhen Dian had completed two rounds of financing, with investors including Daoyuan Capital, Fengrui Capital, Founder Hesheng, Natong Capital, and Suzhou High-Tech.

AI Director DiDi_OK Reveals Workflow for Creating Sci-Fi Short Film 'Candy' Using Seedance 2.5 ↗

Doubao Mobile Assistant Passes First Batch of Terminal-Cloud Collaborative Confidential Computing Evaluations by Telt Laboratory, Achieving L5 Level ↗

Doubao Mobile Assistant passed the first batch of evaluations for terminal-cloud collaborative confidential computing capabilities by China Telt Laboratory, achieving the highest L5 level and covering full-link security for terminal-cloud collaboration. The consumer version was released today and will debut on the Nubia NaviX Ultra, available for purchase on September 16.

Interconnects releases open-source models and an open AI research reading list ↗

How to get up to speed on open models and their implications.

From video to data: how AI is changing multimedia content processing ↗

Checking the site connection security

NVIDIA open-sources OSMO: a YAML orchestrating physical AI training, simulation and robot testing ↗

Robot developers do not have one compute problem. They have 3. A policy is trained on GB200 or H100 clusters, tested in Isaac Sim on RTX GPUs, then validated on a Jetson mounted inside a real robot. Each tier has its own cluster, its own scheduler, and its own glue scripts. NVIDIA OSMO is NVIDIA’s answer to that fragmentation: an open-source, Kubernetes-native workflow orchestrator that lets a team describe the whole pipeline in a single YAML file and run it across every tier without touching infrastructure code.

How Vox Group uses AI technology Aura to solve real-time translation challenges in group travel ↗

Checking the site connection security

OpenAI CEO Altman supports controlled-speed AI development, emphasizing it is not a halt to technological progress ↗

Altman said 'speed control' is not 'halt', emphasizing that safety and technological progress can proceed in parallel. This move triggered a decline in Asian AI concept stocks, with SoftBank's share price plunging 13%. He called for a federal regulatory framework and hoped Washington would assist with international coordination.

JD Jingzao X JoyInside AI Smart Mattress Crowdfunding: 9999 Yuan, Supports 32 Health Monitoring Items ↗

JD Jingzao X JoyInside AI Smart Mattress Launches Crowdfunding Today, as Low as 9699 Yuan After Subsidies, with 1-Year Price Protection + 3-Year Whole-Machine Warranty (10-Year Motor Warranty).

The National Medical Products Administration released the world's first AI + brain-computer interface medical device standard, to be implemented from September 1 next year ↗

The National Medical Products Administration approved and released the world's first brain-computer interface medical device standard that uses AI to process EEG data, to be implemented on September 1 next year. The new standard regulates full-process technical requirements for data, provides a unified quality benchmark for the industry, and reaches internationally leading level.