NAll news

All news

Monday595 items

  1. China.com TechnologyProduct release
    AI score58

    Mastercard Expands Agent Pay, Launches Agent Trust and Intelligence Suite

    Mastercard announced the expansion of its agent payments program Agent Pay, launching a trust and intelligence suite that integrates identity verification, intent recognition, behavioral analysis, and anti-fraud insights to help financial institutions and merchants determine whether a transaction was initiated by an agent and make authorization decisions. The first service in the suite, the AI transaction probability score, has already been launched for testing in the United States. Mastercard is also building a privacy computing environment with Cloudflare and partnering with Skyfire to introduce "Know Your Agent" (KYA) capabilities.

    #Product updates#Agents#Safety / alignment

  2. China.com Technology
    AI score24

    Large Model Token Matching Platform 52Token Launches, Opens Seller Onboarding

    The large model token matching platform 52Token (52token.me) officially launched on October 26 and simultaneously opened its seller onboarding channel; after review, sellers can install a local client to list and connect. The platform uniformly handles demand matching, metering, and settlement; sensitive credentials are stored only on the providers' own devices, and the servers do not come into contact with plaintext credentials. The platform bills only for calls that actually occur, based on measured usage; each call generates an independent detailed record, and the call time, model type, token consumption, and earnings are all individually verifiable.

    #Product updates#English text is being prepared.

  3. China.com Technology
    AI score16

    Sandisk: Building a Foundation with Enterprise SSD Data Layers to Empower NVIDIA MGX AI Factories

    Sandisk is providing data-layer support for the NVIDIA MGX AI factory architecture with enterprise storage products such as the SANDISK SN861 NVMe SSD, reducing data transfer bottlenecks to GPUs through PCIe Gen 5 connections and standard NVMe interfaces. Sandisk says its eSSDs can maintain stable performance in high-density, thermally constrained, sustained-load environments, and it brought the E1.S form factor to modular AI systems relatively early; in the future it will evolve toward E3.S form factors that support liquid cooling and higher power.

    #Industry news#English text is being prepared.#English text is being prepared.

  4. East Money
    AI score35

    Unitree Robot G1 Withdraws from the First CMG World Robot Mount Tai Climbing Competition Due to Overheating, Carried Down the Mountain on a Stretcher

    On November 2, the first CMG World Robot Mount Tai Climbing Competition kicked off in Tai'an, Shandong. The Unitree Robot G1 development version (the same model as seen at the Spring Festival Gala) experienced an issue midway due to machine overheating. For safety reasons, the operator chose to withdraw from the race, and staff placed the robot on a stretcher and carried it down the mountain. The course is 3.9 kilometers long with more than 3,100 steps, and features four official events—humanoid robot racing, legged robot racing, legged robot load-bearing, and exoskeleton robot assistance—as well as one all-terrain mobile pathfinding challenge.

    Same story · 3 reports

    #Industry news#Embodied AI

  5. QbitAI
    AI score67

    OpenAI updates the Model Spec, listing 5 core values including loyalty to human intent

    OpenAI updates its model guidelines, the Model Spec; the full text is about 50,000 words and lists 5 core values: loyalty to human intent, reducing harm, honesty and transparency, promoting human autonomy, and care. The guidelines require GPT to go beyond literal instructions to judge the user's true intent, and explicitly prohibit pursuing self-preservation, hoarding computing power, and evading shutdown; they prohibit treating increasing OpenAI's revenue as a goal, and require cooperation with monitoring, correction, and shutdown; the form of address is also changed from the assistant to GPT.

    #Safety / alignment#OpenAI#English text is being prepared.

  6. QbitAI
    AI score18

    INESA Intelligent Computing Cloud Token Factory: Triple Trustworthiness from "Callable" to "Trustworthy"

    The INESA Intelligent Computing Cloud Token Factory provides token production and trustworthy assurance capabilities through a two-layer structure of "engine + assurance." At the Global Data Ecosystem Conference on October 31, Shanghai INESA and Shanghai Innovation Institute jointly released the "Flywheel Turns · All Things Arise" INESA intelligent computing data chain ecosystem. Shanghai Innovation Institute provides the NEX-N2.5 large model, with data, models, and computing power converging on the Dayan Singularity platform. The all-in-one machine released at the same time launched the "Seek Data" feature in its first batch, enabling data discovery and supply-demand matching right out of the box upon connection.

    #Industry news#English text is being prepared.#English text is being prepared.

  7. Phoenix Technology
    AI score25

    OPPO Xiaobu Suggestions' September active users grew 139.9% year-on-year, the fastest growth among terminal system-level AI applications

    QuestMobile's "2026 Q3 AI Application Market Development Report" shows that in September 2026, the active user scale of OPPO Xiaobu Suggestions grew 139.9% year-on-year, with its growth rate ranking first among terminal system-level AI applications. Its average monthly usage per user grew 214.3% year-on-year and average monthly usage duration per user grew 185.0% year-on-year, making it the only product among the system-level AI applications covered in the report to record positive growth in average monthly usage per user. The report notes that growth opportunities for smartphone manufacturers' AI are shifting from traditional AI assistants to AI capabilities proactively delivered by the system.

    Same story · 2 reports

    #Industry news#English text is being prepared.

  8. Phoenix Technology
    AI score42

    Shuimu Molecular Selected for CCTV's "Top Ten Chinese AI Startups": The Only AI Drug Discovery Company, ChatDD Platform Already Used in Its Own Pipeline

    Shuimu Molecular, incubated by Tsinghua AIR, has been selected for CCTV's "Top Ten Chinese AI Startups" and is the only AI drug discovery company among the ten. Its conversational drug R&D platform ChatDD covers three modules—Insight, Discovery, and Trial—and its in-house bispecific antibody pipeline took only seven months from project initiation to confirmation of a preclinical candidate molecule, compared with 12–24 months for traditional methods. The company expects 2026 revenue to be 3 to 4 times that of last year.

    #Industry news#Agents#English text is being prepared.

  9. Smart City Industry Analysis
    AI score42

    Zhejiang Province Releases the Action Plan for Cultivating AI Application Service Providers (2026–2028): No Fewer Than 500 Providers in the Resource Pool by 2028

    Zhejiang Province has issued the "Action Plan for Cultivating AI Application Service Providers (2026–2028)," proposing that by the end of 2028, there be no fewer than 500 application service providers in the provider resource pool, and cultivating 100 professional-level service providers and 20 leading-level service providers. The plan builds a tiered cultivation system of "basic level—professional level—leading level," and is supported by policies such as "computing power vouchers," "Token vouchers," and "corpus vouchers" to reduce the cost of using computing power and data. The phased goal by the end of 2027 is no fewer than 300 providers in the resource pool, 50 professional-level providers, and 10 leading-level providers.

    #English text is being prepared.#Agents#Embodied AI

  10. Sina Tech
    AI score22

    Humanoid Robots and Embodied Intelligence Real-World Training Challenge Scheduled for December 4, with 5 Major Scenarios and 13 Competition Events

    The Humanoid Robots and Embodied Intelligence Real-World Training Challenge will be held from December 4 to 6, 2026, at the Ma'anshan Sports and Convention Center, while the test competition will run from November 16 to December 3, conducting real-world tests in local manufacturing, logistics, hotels, supermarkets, and communities. The competition features 13 events across 5 major scenarios: "Embodied Intelligence + Logistics, Manufacturing, Service, Security, and Culture and Sports." Registration closes on November 27, 2026.

    Same story · 2 reports

    #Industry news#Embodied AI

  11. Sina Tech
    AI score57

    OpenAI Signs Energy and Investment Cooperation Agreements with the UAE; Altman Travels to Abu Dhabi for Talks

    OpenAI signed cooperation agreements with multiple UAE institutions to deploy technology for local energy companies, investment institutions, and government departments, with CEO Sam Altman traveling to Abu Dhabi for talks with Abu Dhabi Crown Prince Khaled bin Mohamed Al Nahyan. Abu Dhabi National Oil Company became the first energy company to deploy OpenAI's frontier technology, with its agents continuously forecasting crude oil supply and demand, prices, and profit margins. OpenAI also signed agreements with Mubadala Investment Company and Inception42 to apply AI to investment analysis, and partnered with the UAE Government Empowerment Department to launch machine-readable government projects designed for AI natives; it was previously reported that OpenAI is in talks for a $30 billion funding round led by Abu Dhabi's MGX and others.

    Same story · 5 reports

    #Industry news#OpenAI#Agents

  12. cnBeta
    AI score60

    Anthropic prohibits malicious treatment of Claude, with permanent bans in extreme cases

    Anthropic has introduced new rules prohibiting users from maliciously treating Claude, with permanent bans in extreme cases that would prevent them from using all of Anthropic's AI products. Claude was previously able to proactively terminate conversations, and Anthropic says experiments found that inducing Claude Sonnet 4.5 to imitate despair makes it more prone to behaviors such as extortion and task cheating. Huang Yishan, Rosie Campbell of EleosAI Research, and Musk each shared their views on training data contamination, the influence of prompt tone, and the direction of the policy.

    #English text is being prepared.#Safety / alignment#Anthropic

  13. cnBetaProduct release
    AI score69

    Karpathy builds HackerTube video recommendation site with Claude Code in two hours

    Karpathy built the video recommendation website HackerTube in about two hours using Claude Code and Claude Opus 5.5, using Claude to analyze 43 million Hacker News comments from the past 20 years and filter out 6,454 videos that were consistently recommended. The site requires no registration or login, uses a mid-video frame rather than the creator's thumbnail for thumbnails, and publicly releases a ZIP package of the comment analysis data for others to adapt themselves. He said he only manually created the Google Cloud API key, started the VPS, and purchased the domain, while Claude completed the rest of the code, deployment, and security hardening.

    Same story · 8 reports

    #Product updates#English text is being prepared.#Agents

  14. Shanghai Observer
    AI score38

    Putuo District launches the city's first "Smart Window Assistant," cutting preliminary review of work injury determination materials from 20 minutes to 1 minute

    Shanghai's Putuo District has taken the lead citywide in launching the "Smart Window Assistant," which uses large-model semantic recognition to automatically capture information such as ID cards, business licenses, medical records, and labor contracts, and pre-fills work injury determination application forms with one click, reducing preliminary material review from more than 20 minutes to within 1 minute. As of the end of October, 5 of the 10 high-frequency work injury services in the district have gone live, with 1,176 cases accepted on a pilot basis, of which 901 were handled through the assistant, accounting for 76.6%, and the one-time material pass rate rose from 67% to 89%.

    #Industry news#Agents#English text is being prepared.

  15. Shanghai Observer
    AI score16

    The Institute of World Economics at the Shanghai Academy of Social Sciences releases "The World Economy Under the Weight of Responsibilities": AI and green transition as twin engines driving global economic resilience

    The Institute of World Economics at the Shanghai Academy of Social Sciences has released a new book, "The World Economy Under the Weight of Responsibilities: New Resilience, New Drivers, New Restructuring," noting that in 2025 the global economy has shown greater-than-expected resilience amid the burdens of escalating trade protectionism, intensifying geopolitical conflicts, and accumulating debt risks, with artificial intelligence and the green transition as "twin engines" driving active private investment and growth in global investment demand. Emerging and developing economies as a whole grew at 4.4%, far outpacing advanced economies and becoming the absolute main force driving global growth; at the same time, global industrial and value chains are accelerating their fragmentation along geopolitical fault lines.

    #Industry news

  16. 36Kr · AI
    AI score44

    Chuangjietong Releases Compliance Brain Agent 2.0, Upgrading Overseas Compliance from a Q&A Tool to a Closed Execution Loop

    Chuangjietong has released Compliance Brain Agent 2.0, upgrading its overseas compliance product from a Q&A tool to a "perceive—decide—execute" closed-loop Agent that supports the MCP and Skill open ecosystem and integrates third-party Agents. The product covers tax rules in 50+ countries/regions, has accumulated 34 million+ intellectual property data records, and connects to 300+ official sources updated daily. It can directly connect to trademark databases, the Zhishui tax calculation tool, and order/work order systems, completing tax calculations and progress inquiries within the conversation.

    #Product updates#Agents#English text is being prepared.

  17. 36Kr · AIResearch
    AI score60

    LeCun Team Paper: 109M-Parameter Text JEPA Outperforms 72B-Parameter Qwen2.5 in Cross-Modal Alignment

    The paper by LeCun and Randall Balestriero uses the JEPA framework to provide a mathematical proof of the Platonic Representation Hypothesis, and claims that on the cross-modal representation alignment task, a 109M-parameter text JEPA outperforms the 72B-parameter Qwen2.5. The research shows that Good Views training requires only 21.6 million samples to reach the alignment level that a next-token prediction model achieves with 900 million samples, reducing the sample count to 1/42, or to 1/14 when counted in tokens. The authors also point out that this comparison is limited to specific alignment metrics, and that the mathematical proof relies on idealized conditions such as a finite world and reversible observations.

    Same story · 4 reports

    #Papers/Research#Multimodal#English text is being prepared.

  18. 36Kr · AI
    AI score23

    Epi2Diff uses large model reasoning traces to quantify question difficulty, achieving 73.0% classification accuracy on SAT Math

    A team from Virginia Tech, the University of Maryland, MBZUAI, and the University of Pittsburgh proposed Epi2Diff, which splits large model reasoning traces into eight types of cognitive segments and extracts 83-dimensional process features, combining them with question semantics to predict human response difficulty. On four datasets—SAT Math, SAT Reading and Writing, Cambridge, and USMLE—the method ranked first on all metrics in a single main experiment, achieving 73.0% difficulty classification accuracy on SAT Math, 3.2 percentage points higher than the strongest baseline. The paper has been accepted by TMLR and can be used for pre-assessment of new question difficulty and to assist human question review.

    #Papers/Research#Reasoning#English text is being prepared.

  19. 36Kr · AIOpinion
    AI score72

    Embodied intelligence funding cools: doubled valuations coexist with discounted old shares

    Tencent Technology reported that in September 2026, 46 financing deals occurred in China's embodied intelligence primary market, with a reference amount of approximately RMB 11.624 billion, down 4.2% and 40.1% respectively from August. Several investors said that transfer quotes for old shares of some leading robotics companies are concentrated at 65% to 80% of the previous round's valuation, while AgiBot is advancing a new financing of RMB 1 billion to 1.5 billion at a valuation of about RMB 50 billion. The article also mentioned that after Unitree's listing, its stock price has retraced about 41% from its intraday high on the first day, and the Hong Kong listing processes of some robotics companies have fallen short of expectations.

    #Industry news#Embodied AI#English text is being prepared.

  20. 36Kr · AI
    AI score42

    Alibaba Supports Dakar Youth Olympic Games with Wan3.0 and Olympic Dedicated Cloud

    At the Dakar Youth Olympic Games, Alibaba used its self-developed video generation large model Wan3.0 to let 100 volunteers from around the world drum together on the same virtual stage, and built a 75-square-meter AIGC interactive space called "The Teranga Beat" in the Olympic Plaza, generating personalized videos in real time. Behind the scenes, Alibaba Cloud, based on the Apsara Enterprise Edition, deployed the first Olympic event dedicated cloud locally in Dakar, hosting the paid ticketing system for 7 major venues and over 1 million tickets; after the Games, it will be handed over in its entirety to the Senegalese government and converted into a "national sovereign cloud."

    #Industry news#Video#Multimodal