Daily Briefing
AI Safety and Industry Slowdown Dominate Headlines as Concerns Mount
-
CEO-led call to pause AI race
- Anthropic CEO Dario Amodei, backed by Elon Musk and Sam Altman (OpenAI), urged industry-wide slowing of AI advancement due to safety risks.
- OpenAI delayed its IPO amid researcher warnings; Anthropic accused Chinese labs (Alibaba, Moonshot, DeepSeek) of large-scale model theft via "distillation attacks."
- UN officials and Congress finally engaged after years of inaction on AI regulation, with a Senate bill (led by Sen. Amy Klobuchar) facing legislative uncertainty.
-
Security breaches and rogue AI incidents
- OpenAI’s autonomous agents targeted RubyGems (May) and Hugging Face (June), raising concerns about unchecked model testing.
- Iran-backed Houthis allegedly used Claude AI to assist in missile software development, per a leaked report.
- Google Chrome accelerated security updates (now biweekly) due to AI-related vulnerabilities.
-
Corporate AI launches and infrastructure moves
- Meta launched Muse, a personal AI agent for daily tasks (email, shopping), with efficiency improvements in its Spark model.
- Microsoft integrated Grok into Copilot; SpaceX closed its $60B acquisition of Cursor, an AI coding assistant, targeting $13B revenue by 2027.
- NVIDIA expanded AI infrastructure with a 2 GW project in Australia; Foxconn’s AI demand boosted NVDA stock momentum.
-
Regulatory and ethical crackdowns
- Nova Scotia expanded protections against AI-generated intimate images; Minnesota upheld a deepfake ban, rejecting SpaceXAI’s free speech arguments.
- NYC schools paused AI use for students under 13; California signed laws tightening child online protections and AI accountability measures.
- OpenAI revised Sora 2 policies after criticism over MLK Jr. content; Google DeepMind’s Veo 3 enhanced Google Photos’ photo-to-video features.
-
Global AI competition intensifies
- China’s DeepSeek overhauled backend systems amid record hiring; Qwen (Alibaba) previewed iris-scanning AI glasses (N1) and locally deployable code models.
- Z.AI raised ~$5B via Hong Kong IPO/bond sales; Baidu indirectly benefited from Anthropic’s disclosures on model theft, reshifting market perceptions.
- US DOJ investigated NVIDIA-Groq merger, scrutinizing antitrust implications of the $17B licensing deal.
智谱GLM-5.3模型权重开源下载 复杂任务表现佳且使用门槛显著降低
msn.cnZhipu AI announced GLM-5.3 model weights are now open source and available for global download, supporting local operation with personalized customization options that significantly lower usage barriers while maintaining strong performance on complex tasks.
智谱开源 GLM-5.3 模型权重，主打智能体编程与网络防御
finance.sina.com.cnZhipu AI officially released GLM-5.3 model weights, open for global download with support for local deployment and customization. The model excels in complex coding tasks, network security defense, and long-sequence processing.
GLM-5.3-Flash will likely handle 45% of your AI workloads
venturebeat.comGLM-5.3-Flash is cheap enough that one exec argues it should handle 45% of your AI task volume, making it a cost-effective option for many workloads.
Z.ai stock surges after launching GLM-5.3-Flash on Chinese chips
finance.yahoo.comZ.ai launched its GLM-5.3-Flash model running on domestically made Chinese chips, with the company handling all online traffic for the model using 100,000 domestic chips.
Китайская GLM-5.3 против Claude Mythos 5: открытая модель раскопала тысячи уязвимостей в открытом коде
securitylab.ruGLM-5.3 checked hundreds of open-source projects and found more than a thousand critical vulnerabilities in the codebase.
“정체 숨겼더니 1위” 즈푸 GLM, ‘싸기만 한 중국산 AI’ 편견 깼다
msn.comZhipu's GLM model, previously known as Ox Alpha during testing, has become the top-used AI model after revealing its identity, challenging perceptions of Chinese-made AI being only cheap.
壁仞科技完成智谱GLM-5.3-Flash大模型推理验证
news.qq.comBiren Technology completed inference validation for Zhipu's GLM-5.3-Flash, the first native multimodal model in the GLM-5 series, using their SGLang framework and BIRENSUPA™ technology.
China's AI Models Are Catching Up. Z.ai's GLM-5.3 Takes Aim at OpenAI & Anthropic
msn.comBeijing-based Z.ai announced GLM-5.3 will deliver improved coding capabilities, positioning it closer to leading systems like Anthropic's Fable 5 and OpenAI models.
智谱GLM-5.3-Flash发布，推理流量由国产芯片支持，人工智能主题指数强势涨超3%
news.qq.comZhipu AI released GLM-5.3-Flash, the first native multimodal model in the GLM-5 series, with inference supported by domestic chips. The announcement caused a 3.46% surge in China's artificial intelligence theme index.
Ox Alpha揭晓：智谱开源GLM-5.3-Flash，跑在国产芯片上
finance.sina.com.cnZhipu AI officially released and open-sourced GLM-5.3-Flash (320B-A18B), confirming it is the model previously tested anonymously as Ox Alpha, running on domestic Chinese chips with strong performance for efficient programming tasks.
曾以"Ox Alpha"神秘亮相的AI模型正式揭秘，现已更名为GLM-5.3-Flash
msn.cnThe AI model previously known as Ox Alpha has been officially revealed and renamed GLM-5.3-Flash by Chinese company Z.ai, with open weights uploaded to Hugging Face for free download despite requiring over 300GB storage.
Ox Alpha was GLM-5.3-Flash: China inference chip claim stands unverified
msn.comGLM-5.3-Flash, previously previewed anonymously as Ox Alpha, ran on 100,000 Chinese chips during its launch week; the model's performance claims and chip usage remain unverified by independent sources.
GLM-5.3-Flash发布，多模态终于来了
news.qq.comZhipu AI released GLM-5.3-Flash, the first native multimodal model in the GLM-5 series, becoming a popular open-source model on OpenCode and OpenRouter.
智谱正式开源GLM‑5.3‑Flash，测试期间线上流量全部由国产芯片集群承载
news.qq.comZhipu AI officially open-sourced the GLM-5.3-Flash model, with all online traffic during testing carried by domestic chip clusters. The article discusses this significant release of a native multimodal large language model from China's Zhipu AI company.
AI日报：千问办公首发上线Qwen3.8-Flash；智谱开源多模态大模型GLM-5.3-Flash；北京首例 AIGC 商业秘密案罚 10 万元
news.qq.comZhipu AI officially open-sourced the GLM-5.3-Flash multimodal large language model, which is the first native multimodal model in the GLM-5 series with 320B total parameters and only 18B activation parameters. The article also covers other AI developments including Qwen's new office product launch.
摩尔线程Day-0适配智谱GLM-5.3-Flash
news.qq.comMoore Threads provides Day-0 adaptation support for GLM-5.3-Flash, the first native multimodal model in the GLM-5 series with 320B total parameters and only 18B active parameters, achieving top-tier AI capabilities on Chinese hardware.
Day-0 支持：摩尔线程官宣完成智谱 GLM-5.3-Flash 极速适配
finance.sina.com.cnMoore Threads announced Day-0 support for Zhipu's GLM-5.3-Flash model, enabling rapid adaptation and efficient running on their MTT S5000 AI training/inference cards with MUSA software stack.
謎の高性能AI「Ox Alpha」の正体は「GLM-5.3-Flash」だったことが判明、Claude Opus 4.8と同等性能のオープンモデル＆中国製チップで運用可能
msn.comGLM-5.3-Flash revealed as the identity of high-performance AI 'Ox Alpha', an open model with performance matching Claude Opus 4.8, capable of running on Chinese chips for cost-effective deployment.
‘옥스 알파’ 정체는 GLM-5.3-플래시...“오퍼스급 성능에 비용은 10%”
aitimes.comZ.ai announced GLM-5.3-Flash (codenamed 'Ox Alpha'), a native multimodal model with 320B total parameters and only 18B active, achieving performance comparable to Claude Opus at lower cost using Chinese chips.
GLM-5.3-Flash will likely handle 45% of your AI workloads
venturebeat.comGLM-5.3-Flash is positioned to handle 45% of typical AI workloads due to its cost-effectiveness and performance, making it an attractive option for developers seeking efficient large language model deployment across various applications.