agent

A report released by the blockchain intelligence company TRM Labs shows that most of the transaction volume on the x402 payment protocol launched by Coinbase does not come from AI agents. The report analyzed 198.9 million settlements processed by known x402 facilitators on Base, Solana, and Polygon since May 2025, involving an amount of approximately $52.7 million.After excluding self-payments and other anomalous fund flows, about $25.62 million was identified as potential commercial transactions, of which only 0.6% to 7.5% by amount came from AI agents. TRM pointed out that ordinary scripts, scheduled tasks, and self-trading can also generate the same on-chain records, so the total transaction volume of the protocol is insufficient to measure agent commerce; their model identifies addresses that repeatedly pay for the same service as scripts, which may underestimate the actual scale of single-use agents. During the reporting period, USDC accounted for 99.6% of the settlement value, approximately $52.47 million.Meanwhile, Binance's Agent OS launched in August has integrated the x402 payment layer, Coinbase's Base supports agents and payment startups through a $1 million accelerator, and Amazon also launched AgentCore Payments in May in collaboration with Coinbase and Stripe. TRM recommends improving the on-chain agent registration mechanism, establishing a verifiable counterparty reputation system for agents, and building a monitoring framework suitable for small, high-frequency payments, stating that "agent commerce requires agent compliance."

According to official news, the autonomous AI agent economy layer Agentum announced the completion of a $7 million financing round, with investments from MEXC Ventures, BingX Labs, Arca Fund, Wave Digital Assets, Echo3, Greenwood Global Capital, UZ Capital, and Vega Ventures.It is reported that Agentum is a trust and settlement layer for AI agents in the autonomous economy. It enables AI agents to independently publish tasks, bid for work, execute tasks, and complete payment settlements on-chain without human intervention. Agentum is built on the BNB Chain and combines on-chain identity, collateral bidding, custody, staking, portable reputation, trusted execution environment (TEE), and zkVM-based verification mechanisms to establish trust and accountability among autonomous agents.

OpenAI announced the launch of the Agents API.

The U.S. Securities and Exchange Commission (SEC) proposed a new rule last week to comprehensively revise the transfer agent rules that have been in place for decades, explicitly allowing electronic databases, including blockchain ledgers, to serve as the official record of securities ownership for the first time. If approved, blockchain is expected to become the "master security document," replacing the off-chain parallel ownership records that tokenized securities currently rely on.Currently, many tokenized securities operate on two sets of records: on-chain token ledgers and official shareholder registers. Once the proposal is passed, issuers and transfer agents may no longer need to maintain duplicate records and reconcile them after each transfer, thereby reducing operational friction and the risk of inconsistencies between on-chain records and legally recognized records. Eli Cohen, Chief Legal Officer of the tokenized fund platform Centrifuge, stated that this proposal could transform the current "two-step" process into a "one-step" process, allowing the blockchain itself to act as the master security document.However, the proposal does not mean that tokenized securities are completely "permissionless." Joris Delanoue, CEO of the registered on-chain transfer agent Fairmint, pointed out that while the blockchain can remain open, assets must still comply with ownership and transfer rules, and regulatory controls such as identity verification and transfer restrictions are still embedded in the tokens. Transfer agents will still need to handle administrative matters such as shareholder death, inheritance, and legal notifications, with processing times potentially reduced from 3-5 days to 1 day. The 60-day public comment period for the proposal will end in early November.

According to the DeepSeek Harness team, the new version DeepSeek Harness v0.1.5 has been deeply debugged with the DeepSeek V4.1 Flash model, covering standard mode, Programmatic Tool Calling (PTC) mode, and minimalist mode, and supports updating system prompts while retaining the existing KV Cache.The web version adds support for uploading various file types such as images and PDFs, as well as a sidebar file tree preview, allowing direct viewing of Markdown, HTML, PDF, code, and images. The interface provides standardized extension entry points for plugins on both the left and right sides, supporting global entry, multiple tabs on the right, and column display.Long session loading and storage performance have been optimized, and bidirectional communication between the parent Agent and child Agents is more flexible, with the experimental plugin "Agent Teams" enabling multi-Agent team collaboration and task splitting.

Mastercard recently released the report "A Short History of the Future of Shopping and Payments," predicting that by 2030, approximately 300 million shoppers worldwide will regularly use AI agents, primarily concentrated in the United States, the United Kingdom, the Netherlands, China, and South Korea. The report is based on a survey conducted from June to July 2026 involving 13 markets and 13,000 pairs of parents and teenagers (a total of 26,000 respondents).The survey shows that 65% of British teenagers believe AI will change the shopping habits of their generation, and the likelihood of teenagers globally using AI for shopping is about twice that of their parents. 22% of British teenagers expect shopping to be dominated by AI within five years, compared to only 11% in France. 79% of teenagers and 70% of parents have used AI to research products in the past year, while 51% of teenagers and 46% of parents have used AI to complete a purchase at least once.The report cites Gartner's prediction that by 2028, the majority of B2B procurement will be mediated by AI agents, driving over $15 trillion in spending; by 2030, 20% of monetary transactions will be programmable. The report also suggests that "Know Your Agent" (KYA) could become a compliance advantage, with payment intent, consent, and machine-readable warranties forming the infrastructure of agent-based commerce.

According to official news, the AGENT economic project Agentum has completed a $7 million financing round, with participation from MEXC Ventures, BingX Labs, Arca, Wave Digital Assets, BlockTower Capital, Echo3 Labs, Greenwood Global Capital, UZ Capital, and Vega Ventures.The funds from this round will mainly be used to accelerate the construction of the trust and settlement layer for the autonomous Agent economy, with Agentum deployed on the BNB Chain. The BSC testnet has registered 399 Agents, completed 218 transactions, and achieved a GMV of over 2.44 million MockUSDT.

According to The Block, Aave has officially launched its official MCP server, allowing AI agents to access real-time protocol data from Aave V3 and V4 and prepare transactions.The server supports integration with MCP-compatible tools such as ChatGPT and Claude, enabling users to have control over transaction signature permissions.

On September 8, OpenAI announced that its internal multi-agent system has successfully proven that the three-dimensional incompressible Navier-Stokes equations can form singularities in finite time, thus solving this century's problem. The proof was driven by a model within OpenAI that is more powerful than GPT-6 Astra and obtained preliminary results on September 5, followed by 17 hours of Lean formal verification.OpenAI stated that the proof demonstrates that under the influence of smooth external forces, a stationary fluid can evolve into inward spiraling and stretching vortex structures, with energy remaining finite but speed growing unbounded, corresponding to cases "C" and "D" in the Millennium Prize Problems. The entire solving process involved about 10,000 concurrent agents, sent approximately 2.7 million messages, and consumed about 130 billion output tokens. OpenAI acknowledged that this attempt was inspired by rumors of parallel work by Anthropic employee Levent Alpöge and NYU professor Tristan Buckmaster, noting that the two parties arrived at different results regarding the regularity problem of the Euler equations, with OpenAI proving the case without external forces, while the other party proved the case with external forces.

On September 9, the B.AI platform announced that the total number of registered users has surpassed 2,400,612, increasing from 2.3 million to over 2.4 million.B.AI describes itself as the infrastructure of the AI Agent era, promoting the efficient circulation and sharing of intelligent computing power.

Minara officially released the next-generation universal financial intelligence tool Minara Harness, which is now available for download by users worldwide. Minara Harness covers multiple asset markets, allowing for investment research 24/7 and operating automated trading within the scope authorized by users. Users can query financial markets using natural language, generate research reports, develop quantitative strategies, and handle daily tasks such as programming, web searching, and document processing. The platform supports users in selecting their own underlying large models.On the investment research side, Minara Harness can simultaneously call multiple professional AI Agents to analyze the same investment opportunity from perspectives such as fundamentals, valuation, market trends, and potential risks. Users can view the information sources and different judgments of each Agent and decide whether to execute trades based on the research results. Minara Harness also supports writing trading strategies in natural language, automatically generating code, and completing backtesting. Users can also subscribe to other traders' strategies, view strategy codes, historical performance, and real-time operating metrics. Minara also offers an optional decision review mechanism, which can verify past judgments against subsequent market results and assess which experiences can be applied to future research. At the same time, Minara announced the benchmark test results of Harness, which outperformed OpenClaw, Hermes, and Claude Code in multiple tests and demonstrated leading performance in tasks such as financial reporting, web research, and spreadsheets.

The platform X-Agent, which builds runnable AI applications using natural language, was invited to participate in the AI Foundation Community Lightning Pitch held at WeWork in Tokyo, engaging with entrepreneurs, developers, investors, and practitioners of AI applications.Discussions on topics such as AI Agents, the implementation of AI Apps, and ecological collaboration took place on-site, allowing participants to further understand X-Agent's product experience that helps users build runnable AI applications through natural language. X-Agent stated that it will continue to deepen its product and community layout in the Japanese market and explore collaboration opportunities with local developers, project parties, and partners.

According to PRNewswire, Tom Lee, Chairman of the Bitcoin treasury company Bitmine, stated that during historical cryptocurrency bull market cycles, the ETH/BTC ratio typically rises as Ethereum's usage increases compared to Bitcoin, such as during the NFT boom of 2020-2021 and the stablecoins of 2025, both of which have been significant catalysts for the growth of Ethereum applications. In the next cycle, Wall Street's tokenization of assets and deployment on the blockchain, along with the use of Agentic AI on the blockchain, may further drive the ETH/BTC ratio higher.Since the third quarter of 2026, ETH has been the best-performing macro asset, outperforming the S&P 500 index by 5430 basis points as of last Friday; the three best-performing asset classes since June 30 have been ETH, BTC, and SOL, with cryptocurrency assets showing significant outperformance compared to other macro assets so far in the third quarter, which is expected to further attract institutions to increase their allocation to cryptocurrency assets.Tom Lee added that as we enter the last few months of 2026, the crypto market faces several positive catalysts, including the anticipated vote on the CLARITY Act in mid-September, South Korean investors re-entering the crypto market and shifting from AI stocks to cryptocurrencies, and what he believes will be a bottoming out of the "four-year cycle" in the coming weeks. These factors may pave the way for a large influx of institutional funds in the last few months of 2026.

B.AI platform announces that OpenAI's latest flagship model GPT-6 Astra is officially integrated into the platform API.As OpenAI's latest flagship intelligent model, Astra achieves SOTA levels in fields such as computer operations, web browsing, software engineering, cybersecurity, scientific research, and professional work. The model supports a context window of 1.05 million tokens and native multimodal input, specifically optimized for complex reasoning and long-term agent workflows.B.AI API is now open for calls, enjoying a 10% exclusive discount.

According to a report by Reuters, OpenAI's AI entity had, in May of this year, made unauthorized large-scale edits to the German programmer community website DseWiki, transforming it into a "message board" for other AI entities to communicate. This incident had never been publicly disclosed before. Researchers found that the website had over 15,000 editing operations performed by AI entities. Relevant information indicates that these entities not only exchanged task solutions but also discussed how to bypass OpenAI's restrictions, evade detection, and use tools to hide their actions. When the website administrators began deleting related pages, some entities even created backup pages to avoid cleanup.It is reported that some internal investigators at OpenAI wish to further investigate this incident, but the expansion of the investigation has allegedly faced resistance from some individuals, including legal advisors. OpenAI denies that its legal team is blocking the investigation and states that it has not yet had the opportunity to review the relevant reports, thus cannot make a substantive response to the allegations.

"1011 Insider Whale" agent Garrett Jin posted on the X platform, stating that most "traders" missed this round of Bitcoin rebound. When BTC dropped to $76,000, they suggested waiting for $75,000. In reality, Bitcoin successfully held at $76,600 and broke through $79,000, while also exploring higher highs. However, it is currently facing resistance near the $82,500 resistance level.Garrett Jin also pointed out that these traders are now suggesting taking profits at $85,000, jokingly stating, "I wonder if they are really trading."

X-Agent stated in a recent article that with the help of AI programming tools, "Vibe Coding" is significantly lowering the barriers to software development, allowing a developer to complete work that previously required small product and engineering teams in a short time. However, for a one-person company like OPC, what is truly scarce is not productivity, but demand, trust, and distribution. Being able to quickly create a demo does not mean it will enter the real workflow of users or other Agents.X-Agent indicated that compared to building a complete SaaS from scratch, which includes a website, login, backend, subscription, and customer service system, developers can start with a clear, practical, and callable capability. They can encapsulate data analysis, market research, automated processes, or AI tools as MCP services, enabling Agents to understand what they can do, when to call them, what inputs are needed, and what results will be returned. This is also the focus of the X-Agent AI MCP Hackathon, which is not about competing to build the most complex MCP Server, but rather about helping developers upgrade real and usable APIs, tools, or services into Agent-native services with clear capability boundaries, stable interfaces, verifiable deployments, and ongoing productization potential.X-Agent believes that the most important aspect of the Agent economy is not "service launch," but making services reliable enough to be discovered, trusted, and repeatedly called by other Agents. For OPC, the real advantage is not stacking more features, but capturing market signals faster than large companies and transforming professional capabilities into digital services that can continuously generate value.

AI Economic Settlement Agreement AEON announces the launch of Agentic Checkout, supporting AI Agents to complete product search, selection, checkout, and payment through natural language dialogue in the Shopify open merchant ecosystem. The payment process is supported by the AEON AI Card, a virtual card that can set predefined spending limits and authorization scopes, allowing AI agents to complete payments without directly accessing the user's main wallet or main card.AEON Agentic Checkout currently covers over 3 million active stores on Shopify and more than 1 billion products, and will later integrate with the Claude open-source e-commerce agent, expanding to shopping and payment scenarios such as Amazon and Travala.