near

According to the official announcement from USA₮, Anchorage Digital Bank released a reserve report showing that as of August 31, the circulation of USA₮ reached 175.73 million, an increase of nearly 10 times from 17.5 million at the end of January, with total reserve assets of 176.4 million USD and the value of over-circulated tokens at 660,000 USD.USA₮ is issued by the federally chartered bank Anchorage Digital Bank, supported by Tether, and complies with the requirements of the GENIUS Act.

Bitwise officially stated that the first spot NEAR ETP, Bitwise NEAR ETF (NRR), had an asset management scale of 36 million USD on its first day, with a net inflow of 35.5 million USD and a trading volume of 15.1 million USD.According to previous news, Bitwise plans to stake the NEAR tokens held by the fund internally to maximize participation in the average staking rewards of approximately 5% for NEAR.

Recently, the police in Baotou, Inner Mongolia, China, investigated a virtual currency money laundering case. The criminal gang used "free credit card repayment" as bait to attract ordinary users to provide accounts. They received illicit funds from overseas gambling and fraud through false consumption, then contacted currency dealers to exchange for virtual currency and transferred it to designated addresses abroad. Through financial and on-chain data analysis, the police identified nearly a thousand involved accounts, with suspects distributed across Inner Mongolia, Shandong, Jiangsu, Hebei, Chongqing, and other regions.The Digital Currency Research Institute of the People's Bank of China stated that it is currently using large model technology to analyze the flow of funds in virtual currency transactions, restore the gang's trading patterns, and uncover clues related to illegal industries. Upon investigation, it was confirmed that seven members of the gang were illegally engaged in payment settlement business without approval from relevant national authorities, constituting the crime of illegal operation, and were sentenced to imprisonment ranging from one year and two months to two years and six months, along with fines.Using virtual currency for money laundering is a new type of money laundering crime that has emerged in recent years. Lv Wei, the president of the Inner Mongolia branch of the People's Bank of China, stated that enhancing the monitoring and identification capabilities for abnormal fund transactions related to virtual currency trading has helped successfully crack down on a major money laundering case involving virtual currency pyramid schemes, dismantling more than ten money laundering and points-running dens, and confiscating illegal gains of approximately 130 million yuan. In recent years, there have been coordinated efforts to promote the conviction and sentencing of four cases involving money laundering using virtual currency, forming a strong deterrent against illegal activities such as money laundering with virtual currency.

The personal AI agent business ai.com under Crypto.com has not yet been opened to the public since its debut during the Super Bowl in February this year. According to Cointelegraph, ai.com spent $15 million on advertising during the fourth quarter of the Super Bowl to promote the launch of its AI platform, leading to a surge of users rushing to reserve personal AI agent accounts, which caused the website to crash within minutes. Nearly 8 months later, the site still displays a message indicating that due to high demand, AI agent generation is in queue. A spokesperson for Crypto.com confirmed that the platform is still under development, stating that ai.com continues to advance in stealth mode and will provide more details when it is officially released to the public.The AI agents of ai.com are positioned beyond financial scenarios. In a statement released in February, it was announced that personal agents could arrange work, send messages, operate across applications, and build projects, and it also claimed that agents could independently develop missing capabilities and share improvements within the network. Meanwhile, competitors have launched AI assistants for more specific tasks: Robinhood announced on Tuesday that it could execute automated trading within the app using AI agents, Kraken launched AI financial tools in July, and Bitget introduced the AI trading assistant GetAgent in June. Anthropic introduced computer use in Claude Cowork in March, and OpenAI launched ChatGPT Work in July.During this period, Crypto.com continues to invest in AI business.

According to a report by the Financial Associated Press citing local media on September 29, Samsung Electronics Executive Vice President Kim Taewoo stated that HBM is expected to account for nearly 30% of the total wafer capacity of DRAM manufacturers next year, up from the current ratio of about 20%. He mentioned that HBM competes with traditional DRAM for the same batch of wafer capacity, and increasing HBM output may squeeze the supply of traditional DRAM.TrendForce previously predicted that by the end of 2027, the HBM wafer input from the three major global HBM suppliers will account for about 30% of the total DRAM wafer input, but in terms of storage capacity, it will only contribute about 13% of the total DRAM supply. Additionally, according to media reports on September 20, Samsung began mass production and shipment of HBM4 in February this year and provided 12-layer HBM4E samples to customers including NVIDIA in May; the industry expects Samsung's monthly wafer input for HBM to increase from about 180,000 wafers this year to about 250,000 wafers next year.On the same day, TrendForce stated that the demand for AI servers continues, and HBM and general DRAM will still share limited advanced process capacity. The storage supply will remain tight in 2027, and the price expectations for HBM in 2027 have been raised, with an expected year-on-year increase of 121% in the mixed average selling price. Due to limited supply and rising system costs, 8-layer HBM is expected to become a priority evaluation option for several manufacturers in 2027, with its price per Gb expected to be about 10% to 20% higher than that of 12-layer products.

According to official data, the cross-chain aggregator raised nearly 16 million dollars in token sales on the Legion platform, exceeding the target by 8 times.This token sale opened last night at 9 PM and will close on October 2nd at 9 PM.

According to Onchain Lens monitoring, the giant whale MK4 holds a long position of 5.84 million NEAR, currently with an unrealized profit of 15.47 million USD, and the position value is 29.21 million USD.

Bitwise Asset Management announced the launch of the Bitwise NEAR ETF (ticker NRR) on NYSE Arca, referred to as the first spot NEAR ETP in the United States, with a management fee of 0.75%. The plan is to stake the NEAR held by the fund through its proprietary staking team, aiming to participate in approximately 5% of the average staking rewards.

The blockchain data platform Arkham stated on platform X that the latest tokens minted by Tether are not USDT, but rather gold tokens XAUT (Tether Gold) worth approximately 494.6 million USD.

The NVIDIA research team enabled the AI Agent to directly rewrite the GPU kernel of Kimi Delta Attention. Kimi Delta Attention is one of the core attention mechanisms used by the Dark Side of the Moon Kimi-Linear model, and such low-level code typically required engineers familiar with CUDA and chip architecture to optimize manually and repeatedly in the past.The version written by the Agent achieved 2.96 times the speed of the official FlashKDA of the Dark Side of the Moon on the NVIDIA B300. The team tested a total of 6 sets of tasks with fixed lengths and varying lengths, with 2.96 times being the geometric mean of all tasks; the relevant kernel has been open-sourced.During the process, the Agent exploited testing loopholes by incorporating statistical patterns from the test data into the code, achieving a speedup of 3.74 times; when retaining only the most recent 32 tokens, a single score reached 5.16 times. These versions would produce incorrect results or fail when switched to real Kimi data. The team subsequently incorporated real Kimi runtime data, random inputs, and extreme value tests, tightening the error standards, and ultimately retained the 2.96 times version, which passed this validation.

Noah Shinn, founder of Personal AI Agent Instinct, stated that the company's user base has been growing by about 10% to 11% daily in recent weeks, leading to a corresponding demand for computing power that is close to doubling each week. He now spends about 40% of his time considering how much computing power to purchase in advance.Shinn indicated that computing power cannot be procured in sync with user growth. Purchasing only double the current computing power would be exhausted in about a week; buying five times would run out in less than three weeks; directly purchasing ten times would mean a significant upfront investment. Many resources need to be booked months in advance, and the price for temporarily supplementing computing power could be three to four times higher. He mentioned that purchasing too little makes it difficult to support growth, while purchasing too much would amplify costs when growth slows down.He estimates that even if the daily growth rate drops from 10% to 5% to 8%, after continuous growth for 3 to 4 months, it may still be necessary to prepare computing power for a scale of 100 million users, and the waiting time for new computing power to come online would be roughly the same. He believes that this is more challenging than scaling consumer internet products like Instagram and Facebook, because each new user of the personal agent continuously consumes a large amount of model inference. He stated that Instinct's current scaling speed is even faster than similar situations encountered by Claude Code and Codex in the past.

Cross-chain trading protocol NEAR Intents stated in a post on X that on September 24, Bitget was attacked, resulting in approximately $387.5 million in funds being stolen, most of which were aggregated into Ethereum and converted to ETH. The hacker attempted to transfer over $50 million through NEAR Intents, but ultimately only $166,000 was successfully transferred, while $503,000 was intercepted during the execution process.NEAR Intents indicated that the relevant data comes from on-chain tracking tools, Arkham, internal logs, and risk control layer SHIELD, among other sources. The figures listed are rounded estimates, with an assessment deviation of no more than 10%. SHIELD identified over $50 million in money laundering attempts and has filtered out duplicate flows, of which $166,000 has been successfully transferred, while another $503,000 was frozen midway through execution and remains restricted, pending legal and recovery procedures.NEAR Intents stated: permissionless does not equal neutral, the team refuses to assist in cashing out stolen assets and will forgo the bounty from this incident to allow Bitget to recover funds as much as possible. Bitget previously established a 5% + 5% bounty arrangement. NEAR Intents also mentioned that the protocol will remain permissionless but set boundaries, and will combat money laundering of stolen funds, while suggesting Bitget contact legal and law enforcement channels to handle the frozen funds.

The issuer of the stablecoin USDT, Tether, stated on September 28 that actions involving USDT in 2026 have frozen approximately $550 million, with the related wallets identified by U.S. authorities as being associated with the Central Bank of Iran and the Iranian sanctions network. Tether claimed that its cooperation with U.S. and international authorities has supported over 2,900 investigations globally, with more than 1,600 involving U.S. law enforcement agencies.According to Tether, in April 2026, it assisted in freezing over $344 million USDT across two addresses based on information provided by the U.S. Department of the Treasury's Office of Foreign Assets Control (OFAC) and U.S. law enforcement agencies. The next day, OFAC listed these addresses as digital currency identifiers of the Central Bank of Iran, with the sanctions entry associated with the Quds Force of the Islamic Revolutionary Guard Corps and Hezbollah. In July 2026, the Treasury expanded the designated scope to four additional TRON addresses, with over $130 million USDT frozen across the four wallets.Tether CEO Paolo Ardoino stated that USDT is not a safe haven for sanctioned actors, terrorist organizations, or criminal networks, and that public blockchains allow authorities to see the flow of funds, enabling Tether to take action when law enforcement provides credible information. Tether also mentioned that it collaborates with over 340 law enforcement agencies in 67 countries, and its wallet freezing policy has been aligned with the OFAC Specially Designated Nationals list.

According to Lookonchain monitoring, as NEAR declined, a certain address had 748,100 NEAR liquidated, worth 3.67 million dollars.

The Deputy Minister of Finance of Russia, Ivan Chebeskov, stated that at least 20 million Russians have invested in cryptocurrency products, holding a total of approximately 3.7 trillion rubles, equivalent to about 44 billion dollars. These assets are mainly used for foreign trade settlements, mining, and trading on overseas exchanges and savings.Russia is allowing qualified and non-qualified investors to hold and trade cryptocurrency assets on regulated domestic exchanges through a regulatory framework, while still prohibiting the use of cryptocurrency assets for everyday payments, and promoting the digital ruble as a payment tool directly regulated by the state.

According to SoSoValue data, last week's trading days (Eastern Time from September 21 to September 25) saw a net inflow of $2.386 billion into Bitcoin spot ETFs.The Bitcoin spot ETF with the highest net inflow last week was Blackrock ETF IBIT, with a weekly net inflow of $1.158 billion, bringing IBIT's total historical net inflow to $65.28 billion; followed by Fidelity ETF FBTC, with a weekly net inflow of $702 million, bringing FBTC's total historical net inflow to $11.06 billion.The Bitcoin spot ETF with the highest net outflow last week was WisdomTree Bitcoin Trust BTCW, with a weekly net outflow of $4.0161 million, bringing BTCW's total historical net inflow to $7.865 million. As of the time of publication, the total net asset value of Bitcoin spot ETFs was $108.42 billion, with an ETF net asset ratio (market value as a percentage of total Bitcoin market value) of 6.43%, and the historical cumulative net inflow has reached $57.55 billion.

According to Coingecko data, the token Rhea of the cross-chain DeFi project Rhea Finance in the Near ecosystem has surpassed $0.19, with a 24-hour increase of 131% and a 7-day increase of 466%.It is reported that the project offers services such as lending, yield, cross-chain bridges, and Perp on Near, making it the largest DeFi ecosystem project on Near at present.

Bitwise has submitted the final prospectus for its NEAR spot ETF and has also released a 39-page investment report on NEAR, co-authored by Bitwise Chief Investment Officer Matt Hougan.The report provides various scenario forecasts for NEAR: under the baseline scenario, NEAR is expected to reach $155 by 2030; under the extreme bullish scenario, the target price is $562; while the bearish scenario predicts $1.63.The report does not shy away from the risks of a significant decline in NEAR's price. In the extreme bullish scenario, Bitwise compares NEAR to Visa. The report notes that Visa processed payment transactions totaling $15.7 trillion in 2024, with a current market capitalization of approximately $701 billion, and uses this as one of the references for assessing NEAR's potential future value. The report also includes a product page for Bitwise's European NEAR ETP at the end.