evie

According to the latest data from OpenRouter reported by the Daily Economic News, the total global invocation of AI large models last week (from August 31 to September 6) reached 115 trillion tokens, a month-on-month increase of 1.77%. Among them, the weekly invocation of China's AI large models reached 56.72 trillion tokens, a month-on-month increase of 2.83%, surpassing the United States (16.54 trillion tokens, a month-on-month decrease of 3.1%) for nineteen consecutive weeks, ranking first in the world.Among the top five global invocation volumes, four are Chinese models. Tencent's Hyper Yuan Hy4 preview ranks first with 14.7 trillion tokens, a month-on-month increase of 379%; GPT-5.6 Luna ranks second with 12.9 trillion tokens; Zhiyu GLM-5.3 Flash rises to third with 12.4 trillion tokens; DeepSeek-V4-Flash's official and preview versions rank fourth and fifth, respectively.MiniMax M3 returns to the list after nearly a month, ranking sixth, and has become the foundational base for Saudi HUMAIN's Arabic large model. In addition, Xiaomi MiMo-V2.5 and Gemini 3.7 Flash fell out of the list this week.

According to The Block, multiple cryptocurrency companies are actively lobbying the U.S. Securities and Exchange Commission (SEC) to expedite the ETF review process and allow the submission of confidential draft documents. Among them, Grayscale has explicitly requested that SEC staff commit to a response within 45 days.However, Jane Street and Charles Schwab have expressed reservations about this, voicing concerns over the rushed push for listings and the submission mechanism for confidential documents.

The U.S. Securities and Exchange Commission (SEC) has publicly released its request for comments regarding "Novel ETFs," which may hold crypto assets or adopt unconventional strategies. The disagreement centers on whether the filing documents should remain public before the fund begins trading and how quickly the review process should be. Crypto asset management firm Grayscale and the crypto policy organization Crypto Council for Innovation (CCI) support establishing an optional confidential filing period to reduce competitors from submitting imitation documents.Charles Schwab opposes complete confidentiality and suggests that filing documents should be made public at least 75 days in advance. Grayscale requests that the SEC respond within 45 days, while CCI argues that the confidentiality process should not extend the automatic effectiveness or review period. Venture capital firm Andreessen Horowitz (A16z) supports shortening the review time but emphasizes that the rigor of the review should not be reduced; trading firm Jane Street believes that accelerating the process could lead to a decline in product quality, competitiveness, and liquidity.Currently, there are 174 ETFs related to crypto assets in the U.S., with BlackRock's iShares Bitcoin Trust ETF (IBIT) managing approximately $61 billion in assets, accounting for about 38% of the total assets of related ETFs. The SEC will decide whether to adjust the confidentiality arrangements for filings and the speed of the review.

OpenAI released a video on X, previewing the upcoming release of the Astra model (the video content switches between ASTRA and A6TRA), and the model name may be GPT-6-Astra.

The IPO review status of Yangtze Memory Technologies Co., Ltd. for its initial public offering and listing on the Sci-Tech Innovation Board has been changed to "Inquired.""Inquired" means that Yangtze Memory has officially entered the listing review stage at the Shanghai Stock Exchange, where a round or multiple rounds of written Q&A will be conducted regarding the company's information disclosure, Sci-Tech attributes, and other issues.Yangtze Memory is China's largest 3D NAND flash memory manufacturer and the only 3D NAND original factory. It has independently developed the Xtacking architecture and is a memory IDM company that integrates chip design, production manufacturing, and packaging testing.

OpenAI CEO Altman stated that the voluntary review conducted by the Trump administration on the company's upcoming Astra model is a "productive process," and he indicated that as artificial intelligence capabilities continue to improve, strengthening communication with government officials around the world will become increasingly important. The upcoming Astra model is the company's first to be rated with a "critical" cybersecurity capability level.OpenAI stated that such models require stricter safety measures during development and prior to release. The Trump administration has established a voluntary framework for reviewing certain advanced AI models before their release, which may include allowing government officials access to the model for up to 30 days. Altman confirmed that the Trump administration has reviewed Astra and noted that the process is voluntary in nature. He mentioned that as these models reach quite high levels of capability, close cooperation with security agencies in the U.S., the U.K., and other parts of the world will become increasingly important, and Astra is "an important step forward."

OKX founder Star posted a reminder that depositing from high-risk addresses to OKX may trigger stricter anti-money laundering and risk reviews. Depending on the specific situation, the review may last for 15 days or even longer, and some account functions and funds may be restricted; for accounts confirmed to be involved in high-risk or illegal activities, OKX may terminate services.

Polygon Labs announced that it has completed a SOC 2 Type 1 examination for the Polygon Open Money Stack API and dashboard, covering the security trust services criteria. The independent auditing firm Schellman & Company, LLC determined that Polygon has implemented appropriate security controls.SOC 2 is an auditing framework established by the American Institute of Certified Public Accountants. A Type 1 examination assesses whether the design of the control system is correct at the time of the examination, while a Type 2 examination is ongoing and will test the execution of relevant controls over several months of actual operation. The report states that complementary user entity controls are not necessary for Polygon to fulfill its service commitments.Polygon Labs CEO Marc Boiron stated that as a technology provider handling real funds, it is essential to meet the standards of financial institutions, and most crypto infrastructure has never been audited against this standard for systems actually integrated by customers. Type 1 confirms that the control design is correct, and Type 2 will confirm its daily operation.

On September 2, the B.AI platform announced that Tencent's Hunyuan team has officially launched the API service for the latest open-source flagship model, Hy4 Preview, which is open for developers to access.This model is an early version of the Hy4 series, utilizing a 770B sparse MoE architecture, activating 49B parameters per token, and supporting a 1M ultra-long context window. Developers can directly call it through the official B.AI API channels by logging in at chat.b.ai/chat or visiting docs.b.ai/zh-Hans/llmservice/models/hy4-preview/ for more details.

According to The Wall Street Journal, the U.S. Securities and Exchange Commission (SEC) has intensified its scrutiny of companies behind so-called Special Purpose Vehicles (SPVs). According to informed sources, SEC examiners have been requesting registered investment advisors to provide evidence that their SPVs actually own or hold shares in the private companies they claim.Some insiders have indicated that, ahead of the highly anticipated initial public offerings of SpaceX and Anthropic's planned listing, such funds have ramped up their marketing efforts, while complaints from investors have also increased, prompting regulators to decide to strengthen their scrutiny. According to one of the insiders, the SEC's current reviews typically include document requests, and some reviews also involve on-site interviews. The review process can last anywhere from several weeks to a year.

According to the Ontology blog, its core development team discovered potential security vulnerabilities during routine security checks. Due to a high emphasis on the security and integrity of the mainnet, Ontology has immediately suspended block production on the Ontology mainnet to allow the technical team and validators to conduct a comprehensive security review of the network and related components. Ontology stated that no security incidents have been confirmed at this time, and there are no indications that user assets have been lost or stolen. ONT, ONG, and other on-chain assets are currently assessed to be unaffected; this suspension is a preventive security measure rather than a response to confirmed asset losses or ongoing attacks.During the review period, block production will remain suspended, and on-chain transactions cannot be processed. Users do not need to transfer or take any action regarding their ONT, ONG, and other on-chain assets, but should avoid conducting time-sensitive on-chain transactions until the network officially confirms its restoration. The duration of the suspension has not yet been determined, and the team will prioritize the comprehensiveness of the security review over speed; block production will not resume until the network has been thoroughly assessed and confirmed to be secure. The Ontology team is collaborating with validators and ecosystem partners to address this matter and will update progress through official channels, with a separate announcement to be made before and after the network restoration.

Tencent Hunyuan stated that since the Hunyuan Hy4 preview was first launched on WorkBuddy on August 28, 2026, it has received positive experiences and enthusiastic feedback from a wide range of users and developers due to the significant enhancement of its Agent capabilities. On the first day of launch, there were queuing situations in the WorkBuddy task queue, for which we sincerely apologize.To ensure everyone's user experience, the company has conducted emergency scaling of the Hy4 preview inference cluster and will continue to dynamically allocate resources. However, due to the limitations of high-end computing power and peak concurrent usage, it cannot be ruled out that queuing situations may continue to occur during certain periods.

On August 20, MANTRA Chain released a complete review report of the security incident, confirming that the attacker exploited an unsigned integer underflow vulnerability in the balance accounting layer of the upstream dependency cosmos/evm, unauthorizedly transferring a total of 720,923,967.99 MANTRA from two addresses, valued at approximately 3.6 million dollars based on the price before the attack. Among them, the attacker transferred 600,000,035.56 MANTRA from the on-chain burn address and 120,923,932.44 MANTRA from a genesis-era multi-signature address related to an early incentive program.MANTRA stated that this incident did not involve the leakage of validator keys, administrator privileges, governance control, or multi-signature signers; the attacker did not require privileged access and could complete the attack solely through unauthorized contract deployment and self-funded wallets. The first abnormal transfer occurred at 19:06 UTC on August 20, when the attacker transferred approximately 600 million MANTRA from the burn address; subsequently, at 22:59 UTC, another transfer of approximately 120.9 million MANTRA was made. The chain subsequently stopped operating at 23:13 UTC and resumed after upgrading to v8.4.0. The entire network interruption lasted for 30 hours and 13 minutes.This vulnerability was not an issue with MANTRA's self-developed code but originated from the cosmos/evm module, which is responsible for providing EVM functionality on the Cosmos SDK. The vulnerability allowed the attacker to execute unsigned balance deductions without checking if the balance was sufficient, causing an overflow of values and bypassing normal account authorization logic. MANTRA stated that as of today, no funds have been recovered, with approximately 37.96 million MANTRA (accounting for 5.27% of the total transferred) still remaining in the attacker's address, which has been frozen due to the chain's suspension and v8.4.0 restrictions. The remaining funds have flowed to related trading platforms, and the recovery efforts have entered the law enforcement investigation stage. In the future, monitoring of accounts that cannot normally authorize transfers, burn addresses, and other historically "non-transferable" addresses will be strengthened, and efforts will be made to promote improvements in the security vulnerability disclosure process within the Cosmos ecosystem.

Tencent Hunyuan has released and open-sourced the next-generation large language model Hy4 preview. This model has a total of 770B parameters and 49B active parameters, with a context length exceeding 1M. It demonstrates strong capabilities in real productivity tasks such as coding, office work, and science, firmly placing it in the top tier of open-source models. Hy4 preview significantly expands in model size, context length, and data scale, and enhances real-world performance through high-quality data co-built with Tencent experts in software engineering, gaming, finance, security, and deep collaboration with products like WorkBuddy.In software engineering, the model enhances understanding, planning, debugging, and verification capabilities for long-range development tasks, enabling the construction of complex front-end projects like a Three.js miniature town from scratch. In game development, it supports generating playable prototypes from a single sentence and can complete a full demo in Unity. In smart office applications, it can handle complex financial audits, filtering, analyzing, and delivering from multiple documents. In scientific research, it achieves acceleration in tasks such as molecular dynamics simulations and has initially formed a recursive self-improvement feedback loop.Hy4 preview can be experienced in Tencent products such as WorkBuddy/CodeBuddy domestic and international versions, Yuanbao, ima, and can also be accessed via API calls through Tencent Cloud Tokenhub and OpenRouter. WorkBuddy/CodeBuddy will launch a limited-time free activity for two weeks. Since the reconstruction of the infrastructure, the Hunyuan large model has iterated a major version approximately every two months, continuously optimizing through a preview-first and formal version-following approach.

According to Gate market data, on August 28, both the Hang Seng Index and the Hang Seng Tech Index rose in the morning session. By the midday close, the Hang Seng Index was up 0.47%, and the Hang Seng Tech Index was up 0.6%.Hong Kong stock telecommunications infrastructure concept stocks led the gains, with Junzhi Group and Huanlian Lianxun rising over 4%, and Zhongji Xuchuang rising over 3%. Domestic property stocks collectively rose, with Xuhui Holdings Group and Sunac China rising more than 5%.

According to Bloomberg, the U.S. SEC submitted a new proposal to the Office of Management and Budget (OMB) on August 25, aimed at establishing a new custodial rules framework for investment advisors and investment companies holding client digital assets.The proposal will clarify the custodial framework for crypto assets, addressing previous questions from investment companies and advisors about how to hold digital assets for clients without violating regulatory requirements. It will also eliminate certain existing custodial requirements that the SEC considers outdated due to market evolution and current trading practices.The SEC stated that the proposal is part of Chairman Paul Atkins' efforts to modernize the regulatory framework. The complete details of the proposal will be made public after the OMB completes its review, after which it will be returned to the SEC, possibly with suggested modifications, and then voted on by a committee currently composed of three Republicans, before being announced publicly. As a matter of course, the SEC typically seeks at least 60 days of public comment before finalizing the proposal, followed by another vote for it to take effect.

According to The Block, the rescheduling date for Tornado Cash developer Roman Storm has been postponed to April 26, 2027, because the motion for his release on bail is still under review.Previously, U.S. prosecutors had pushed for the rescheduling to take place in October 2026. Tornado Cash is a mixing protocol in the Ethereum ecosystem, and Roman Storm, as one of the core developers, has faced judicial proceedings due to related charges. This schedule adjustment is directly related to the progress of his motion hearing.

According to Gate market data, the main Hong Kong stock indices fluctuated in the morning session. As of the midday close, the Hang Seng Index rose by 0.68%, and the Hang Seng Tech Index rose by 1.58%. Tech stocks rebounded in the morning, with Xiaomi Group rising over 4%, Meituan and Kuaishou rising over 2%, and Baidu, Alibaba, and Tencent Holdings rising over 1%. The financial sector was also active, with CICC and Shenwan Hongyuan Hong Kong rising over 8%, and Nanhua Futures and China Merchants Securities rising over 5%.

According to LiveMint, Nvidia CEO Jensen Huang stated in the All-In podcast that he personally reviews the compensation and remuneration proposals for the company's approximately 42,000 employees at the end of each cycle, and he almost always increases operating expenses. He said, "If you take good care of your employees, everything else will fall into place," and noted that there are more billionaires born from his management team than from any other CEO.The report cited a survey of about 3,000 Nvidia employees, stating that approximately 76%-78% of employees have become millionaires, with about half having a net worth exceeding $25 million. Employees can purchase stock through the company's stock purchase plan at a 15% discount, and Nvidia's stock price has increased nearly 3,800% since 2019, significantly boosting employee wealth.