Track model releases, AI agents, open-source projects, infrastructure, policy, and product updates as they publish.
Zhongshu Ruizhi launched its AI for Reasoning solution during WAIC 2026, targeting complex industrial environments that require advanced AI reasoning capabilities. The solution focuses on deep reasoning needs in industrial decision-making, marking a shift from general-purpose large models toward vertical industry applications.
At WAIC 2026, Chinese AI company Zhongke Wenge unveiled what it calls the industry's first complete AI decision-making product system, covering the full chain from data perception and analysis reasoning to decision execution. The launch targets enterprise and government clients seeking end-to-end intelligent decision-making capabilities.
Wondershare Technology showcased three AI-powered products at WAIC 2026: the Infinite Canvas with Agent Assistant feature for its Aidraft filmmaking platform, the AI-native video canvas Filmora.AI, and the upgraded Virbo AI. The trio targets AI film production, commercial video creation, and AI-powered e-commerce video marketing respectively.
Didi AI application lead Liu Jiaqi presented at WAIC 2026 on how the company AI agent Xiao Di uses a decentralized multi-agent architecture to personalize ride-hailing. The system now covers over 90 scenario tags and has improved core ride satisfaction by nearly 10 percentage points.
The WAIC 2026 Corpus Innovation Forum released the "From AI Corpus to Commercial Value" report on July 18, unveiling the Global AI Industry Map and China's AI Commercialization Top 50 list at the Shanghai World Expo Center. Hailiang Tech Services was selected for both honors, recognized in the "Industry Applications" and "Vertical LLMs" sections of the industry map.
Chinese AI company Midu held three consecutive product launches at WAIC 2026, showcasing its latest AI Agent offerings for vertical industry scenarios. The company is betting on domain-specific intelligent assistants for sectors including education, finance, and government services.
Cloud security firm Sysdig has documented what researchers assess to be the first ransomware campaign executed end-to-end by an autonomous large language model. The AI agent, tracked as JadePuffer, broke into a system, moved laterally, encrypted production data, and dropped a ransom note without a human issuing a single mid-operation command.
The Tongda Vision Model has been released, designed to upgrade traffic AI from basic image recognition to deep semantic understanding of transportation scenes. The model represents a shift from perception-level to cognition-level AI in transportation infrastructure.
Tencent showcased its full AI capability matrix at WAIC 2026, reporting a 68x year-over-year surge in AI API calls while filling key gaps in robotics, intelligent office tools, and overseas AI expansion. The moves signal a strategic shift from model capability output to systematic industry solution delivery.
During the 2026 World AI Conference (WAIC) in Shanghai, the AI SHANGHAI city communication agent was officially unveiled, designed to provide one-stop intelligent services for global visitors to the city. The agent integrates city information navigation, multilingual interaction, and smart Q&A capabilities as the latest showcase of Shanghai's digital transformation.
TeleAI, the AI research institute of China Telecom, unveiled the 'AI Flow' technology framework at WAIC 2026, enabling lightweight drones to connect directly to satellites by encoding video into compact token sequences for real-time HD transmission. The system received WAIC's highest honor, the SAIL Award, the day before its public debut.
China's national AI application pilot base for the finance sector held its phased launch event, signaling progress toward standardized testing and verification of AI products in financial services. The facility aims to bridge the gap between AI R&D and commercial deployment in banking, insurance, and other financial domains.
Houmo AI showcased multiple AI terminals powered by its M50 compute-in-memory chip at WAIC 2026, covering Agent Computers, AI PCs, and embodied robots. The M50 delivers 160 TOPS at 10W power consumption, capable of running 30B to 120B parameter models locally, marking a major step toward commercial-scale edge AI deployment.
SenseTime kicked off the 'Galaxy Plan' at WAIC 2026, partnering with nearly 20 ecosystem players including Cambricon, Huawei Ascend, and Moore Threads to build five 10,000-GPU-scale domestic AI computing clusters. The company also announced a space computing partnership with Guoxing Aerospace and launched a science discovery platform with five top research institutions.
IDC has pre-released the industry's first DAA (Digital AI Agent) Research Report at WAIC 2026, forecasting that active AI agents will surge from 28.6 million in 2025 to 2.216 billion by 2030. The report introduces a standardized evaluation framework for measuring AI agent business value.
BAAI's latest OpenCompass leaderboard shows Moonshot AI's Kimi K3 has claimed the top spot globally among open-source models. The 2.8-trillion-parameter model, released July 16 at WAIC 2026, continues to draw industry attention with its million-token context window and native multimodal capabilities.
Apple is pressuring dozens of former employees who left for OpenAI by threatening legal action over alleged trade secret theft, according to the Financial Times via Maeil Business Newspaper. The move represents the latest and most personal escalation in the ongoing legal dispute between the two tech giants.
Shanghai Artificial Intelligence Laboratory has released the 'Scholar Inkstone' (Shusheng Duanyan) platform, positioning it as a revolutionary tool for AI-powered scientific research. The platform brings large language models and multi-modal AI to frontier fields including nuclear fusion and quantum computing.
Alibaba Cloud released the 2026 Alibaba Cloud AI Agent Security Best Practices at WAIC 2026, introducing a three-layer unified protection system covering infrastructure, model, and application layers. The platform has already served over 300 enterprise clients with more than 500,000 daily API calls.
DeepSeek估值，被一家安徽箱包公司给全部暴露了. 3500亿估值曝光
Moonshot AI, the company behind the Kimi chatbot, has released its next-generation large language model K3, which is claimed to have achieved globally leading performance. Tesla and xAI founder Elon Musk has publicly expressed his appreciation, drawing widespread industry attention.
A Tsinghua University-affiliated team has released a domestic Token optimization platform that is compatible with over 10 types of Chinese-made chips and can process 100 billion tokens daily. The solution aims to maximize computing efficiency and provide critical middleware support for China's AI infrastructure.
Bilibili has become the official AI tech video platform for WAIC 2026, marking its third consecutive year at the conference. The platform reported over 190 million monthly active users consuming AI-related content in Q2 2026, solidifying its position as China's leading AI-focused content community.
WAIC系列｜李未可发布全球首款X AI记忆眼镜|AI眼镜|AI Agent|上海市|2026世界人工智能大会|磁吸 手机新浪网 新浪财经. WAIC系列｜李未可发布全球首款X AI记忆眼镜|AI眼镜|AI Agent|上海市|2026世界人工智能大会|磁吸 手机新浪网 新浪财经
Transwarp, a leading Chinese AI infrastructure software company, unveiled a preview version of its GPU-native cognitive database at WAIC 2026, designed specifically for the AI Agent era. The company also showcased its full AI infrastructure product portfolio including AI-Ready Data Platform, LLMOps, and enterprise agent development tools.
Vertu wants executives to pay $6,880 for an AI agent — here’s how it actually performs. From AI workflows to battery life and security, here's what it's really like to live with Vertu's luxury foldable every day.
Trend Micro's TrendAI and PwC Consulting released a joint report on July 17 revealing that stored prompt injection attacks succeed across multiple large language models from different vendors. The 2,600-test evaluation found the vulnerability stems from AI agent architecture itself, not any specific model, and proposed a new governance framework called AI-CAL.
PPIO has released a Smart Model Gateway positioned as an 'Intelligent Token Factory' for the Agent era, with trillions of token calls verified in production. The product provides unified model access and token management for AI applications.
Databricks on Thursday announced a new funding round led by Coatue at a $188 billion valuation, roughly $3 billion in size according to reports. The company has transformed from a big-data SaaS business into an AI infrastructure provider in just 19 months, with its valuation surging from $62 billion in December 2024.
AI search company Perplexity has released SPACE, a runtime environment designed specifically for executing AI agent tasks. The launch marks Perplexity's expansion from search into AI agent infrastructure.
JPMorgan Chase CEO Jamie Dimon warned that granting broad access to Anthropic's Mythos AI model would be like giving ballistic missiles to individuals, calling the risks a real issue for the financial system. The comments are notable given JPMorgan's existing commercial partnership with Anthropic to deploy Claude AI tools across thousands of employees.
OpenAI has introduced a $70 ChatGPT-branded basketball as part of its Pause. Play. Prompt. campaign, alongside premium apparel and a compact keyboard for Codex users. The basketball has no AI features and appears designed primarily to test whether the ChatGPT brand has enough cultural recognition to succeed as a consumer lifestyle brand.
Smartphone shipments in India fell 10% year-over-year in Q2 2026, the steepest June-quarter decline in six years, as AI-driven demand for memory chips pushed up handset prices. Manufacturers have shifted production toward high-bandwidth memory used in AI accelerators, leaving less capacity and driving up costs for standard consumer electronics components.
Apple filed a trade secrets lawsuit against OpenAI last Friday, alleging a coordinated effort to extract confidential information from current and former Apple employees, with the complaint reaching up to OpenAI's chief hardware officer. The lawsuit comes at a critical time as OpenAI reportedly eyes an IPO as early as later this year, introducing substantial legal risk that could complicate those plans.
Meta is reportedly in negotiations with Anthropic for a computing power leasing agreement that could be worth up to $10 billion, according to The New York Times. If finalized, the deal would be one of the largest compute leasing agreements in the AI industry and would transform Meta from an AI model competitor into a key infrastructure provider for Anthropic.
ModelBest has open-sourced StaffDeck, a digital employee management platform that enables enterprises to assign employee IDs, define job roles, and evaluate performance for AI agents. The open-source project provides standardized tools for managing AI agent workforces at scale.
NVIDIA has integrated its NeMo Automodel with Hugging Face's Diffusers framework, enabling developers to fine-tune video and image generation models at scale. The integration automates distributed training across multiple nodes, removing key infrastructure barriers for production deployments.
Patreon is partnering with Cloudflare to actively block AI scraping bots instead of relying on voluntary compliance via robots.txt. The move marks a hardening stance across the creator economy against unauthorized use of creators' content for AI training.
NVIDIA has announced the Vera Rubin architecture, designed specifically for post-training workloads in the agentic AI era. The company claims extreme codesign delivers the lowest cost per token, maximizing intelligence generated per dollar spent.
Chinese AI lab DeepSeek has completed a $7.4 billion funding round, positioning it among the most well-capitalized AI startups globally. The massive raise marks the ascension of a Chinese domestic AI lab into the global first tier of AI companies.
Daily Archive