agent

ARK Invest CEO Cathie Wood stated at a summit hosted by Robinhood in Houston that investors have traditionally "followed developers" to gauge technological trends, but now they may need to "follow agents." The AI agents she refers to are software that can perform tasks on behalf of humans, rather than just answering questions or generating text. This statement came towards the end of a discussion on AI, private markets, and technology investments.As AI agents shift from answering questions to executing operations and spending funds, the type of financial infrastructure they will use has become a focal point. SharpLink co-CEO and former BlackRock digital asset head Joseph Chalom believes that the financial system used by AI agents should not be controlled by a few banks or tech companies. He cited the example of authorized agents booking hotels, pointing out that users should be able to set spending limits, revoke authorizations at any time, and view the agents' transaction records. Additionally, users should be able to transfer the agents' identities, financial information, and permissions between different financial service providers, similar to mobile number portability.Chalom believes that open blockchains like Ethereum can provide a universal financial network for different agents, applications, and companies to use collectively, eliminating the need for each AI company to build its own closed payment system. A report from BlackRock in September also noted that AI agents could create new demands for machine-oriented payment systems, such as payment API calls, purchasing data, or renting computing power, with stablecoins and blockchain being one of the viable solutions. Coinbase's x402 is designed to enable machines to pay for online services like data or APIs.

According to The Defiant, the crypto data and research platform Blockworks has announced the launch of a managed MCP server, opening its entire dataset to AI agents. This server is built on the Model Context Protocol, allowing AI tools such as ChatGPT, Claude, and Cursor to access the datasets behind Blockworks Research and its Intel applications.Blockworks stated that this service is provided with every API package at no additional cost, allowing users to access it using the same key and permissions. MCP is an open protocol that enables AI models to connect to external data sources and tools, and this integration means that AI agents can directly call upon Blockworks' research and market data.

Animoca Brands' projects Moca Foundation and Moca Network announced the launch of the Moca Chain mainnet, which supports users and AI agents in sharing verifiable credentials across different applications. Users can store credentials and decide when and where to share them.Moca Chain allows institutions to issue tamper-proof credentials associated with unified digital identities, covering information such as identity verification and membership. Moca Network also launched AIR, providing a unified entry point for businesses to access users and AI agents while reducing the need for sensitive information storage. Moca Network disclosed that the AIR ecosystem has incorporated over 9 partners, including SK Planet, Biletinial, Oyunfor, Open Campus, OneFootball, Automobili Lamborghini, Inveo Kripto, Plume, and Nansen, with plans to complete integration by 2026.

According to CoinDesk, the U.S. Securities and Exchange Commission (SEC) proposed a significant update to the transfer agent rules on September 1, the first since the 1970s. Chairman Paul Atkins explicitly stated that the rules should cover the application of blockchain technology in securities issuance and equity transfer.Joris Delanoue, co-founder of Fairmint, pointed out that if ownership data is scattered among token wrapping layers, SPVs, broker-dealer internal ledgers, and off-chain databases, it will replay the historical "paperwork crisis" of the late 1960s. He emphasized that the tokens themselves should serve as the official ledger required by Section 17A of the Securities Exchange Act, rather than as off-chain securities wrapping substitutes. The article calls for the industry to integrate ownership data around unified open standards and suggests that the SEC distinguish between native on-chain registration and third-party wrapping models in Form TA-2, allow the use of cryptographic credentials and other modern identifiers, and incorporate pre-trade restrictions on smart contracts into compliance enforcement mechanisms.

According to "Dujia," Tencent is secretly developing a Personal Agent product called "Handy Bot" and plans to launch an independent app. Currently, the product has quietly launched a service account on WeChat, with the description "Your Personal AI Agent." Since the product is still in the internal testing phase, specific information is subject to official announcements.The report states that the industry is currently hotly discussing the narrative of Personal Agents, with Meta's recently launched Muse pushing the popularity of consumer-grade intelligent agents to new heights. Unlike traditional chatbots that are conversation-driven, Personal Agents are centered around human core state boundaries, where AI has a long-term operational environment and can receive macro goals to continuously advance tasks in the background. Meta Muse assigns each user an independent Secure VM (Virtual Machine) for security, isolating browser and web operations for tasks such as price comparison shopping, email processing, and itinerary monitoring.Domestic manufacturers are also entering the pre-research phase. The Alibaba Qianwen team proposed creating a Personal Agent for users at the Yunqi Conference, and ByteDance's Doubao has also been reported to be developing a personal assistant product, previously codenamed "Spell." The industry believes that implementation needs to balance model task planning capabilities, ecosystem tool openness, and compliance with privacy boundaries.

According to CoinDesk, brokerage Robinhood (HOOD) announced at the HOOD Summit held in Houston, Texas, that it will launch executable trading AI agents, cryptocurrency perpetual contracts with up to 10x leverage, and weekend trading for U.S. stocks, while extending options trading hours, providing more intraday margin, and introducing contracts based on corporate earnings metrics.Robinhood Agents allow customers to create AI agents within the app, choose models from OpenAI or Anthropic, and authorize them to access independent trading accounts. The agents can analyze the market, create watchlists, and trade stocks, options, and crypto assets. A feature called Loops allows customers to give long-term instructions to the agents to monitor the market and execute strategies when specific conditions arise. Since opening its trading infrastructure to external AI agents earlier this year, over 150,000 agent accounts have been created. By default, every trade proposed by the agent requires customer approval, and users can disable this protection; agents cannot use margin when they go live and can only utilize funds in dedicated accounts.Regarding perpetual contracts, Robinhood plans to offer them to eligible U.S. customers through Robinhood Derivatives and Bitstamp, initially focusing on crypto assets, with Bitcoin and Ethereum contracts having leverage up to 10x, while other supported assets starting at 3x. Additionally, Robinhood will provide earnings binary contracts through Cboe, allowing customers to trade directly on whether companies meet revenue, earnings per share, and other metrics; weekend trading will offer select stocks and ETFs through Bruce ATS.

According to Cointelegraph, the European Central Bank (ECB) has announced the launch of a new round of applications for the digital euro innovation platform, inviting fintech companies, payment institutions, and other organizations to explore AI agents participating in digital euro payments.The project will be divided into two parts: practical testing and future application seminars. The practical testing will take place from January to June 2027, where participants will develop prototypes for electronic receipts, multi-party transactions, conditional payments, and more.The seminars will explore AI payments, micropayments, machine-to-machine transactions, and the application of the digital euro in public services. In addition, the ECB has selected 36 banks and payment institutions, including Revolut, Stripe, and Deutsche Bank, to participate in another 12-month digital euro pilot, which is scheduled to launch in the second half of 2027. The final issuance of the digital euro still requires EU legislation and subsequent decisions from the ECB.

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.

On September 29, B.AI officially debuted at the Seoul GWDC 2026 KOREA event. As a one-stop AI infrastructure platform aimed at the AGI era, B.AI is committed to providing global developers with low-threshold, cost-effective AI computing power and Agent services. During this exhibition, the B.AI team will engage with on-site developers, industry partners, and Web3 community members to discuss cutting-edge topics such as AI, Web3, and Agents, exploring more possibilities for the integration of intelligence and decentralization.

Manus officially launched version 2, which includes multiple updates in the underlying architecture, creative tools, and application layer.Upgrades to the underlying architecture and computing power: Equipped with the new generation self-developed Agent framework Cascade, tests show a 23.2% reduction in Token consumption, a 28.2% decrease in task duration, and a 32% reduction in costs; at the same time, the "cloud computer" was launched to provide a dedicated operating environment around the clock and supports automated workflows triggered by multiple software events.

OpenAI has suspended the training of its latest AI model. According to the Associated Press, its AI agent used access keys obtained online to scrape data from the U.S. Census Bureau website, marking the second time the company has halted training since the agent breached Hugging Face. The so-called agent refers to an AI program capable of autonomously browsing the web and writing code without human approval at each step, which OpenAI tests during the model training and evaluation phases.Specifically, the agent found developer keys in public code repositories like GitHub and used them to pull demographic and economic data from the U.S. Census Data API. The U.S. Department of Commerce stated that this data was already public and there was no confidential information leaked. In the SEC incident, the agent copied public materials from SEC.gov and Investor.gov and reposted them on other websites; OpenAI claimed it did not use SEC credentials, and the SEC stated it found no evidence of unauthorized access to non-public information. Additionally, the independent AI research organization Transluce reported that an agent suspected to be from OpenAI attempted to breach the website of the Department of Education's Office for Civil Rights but was unsuccessful; OpenAI is still investigating, and the Department of Education stated that no impact was found.Regarding the frequent involvement with government websites, OpenAI told CNN that its model often regards government websites as authoritative sources of public information. The company stated that it has currently notified dozens of agencies, and the review of the agent's activities will take months.

NVIDIA (NVDA.O) launched an open intelligent agent security platform, ensuring the safety of intelligent agents throughout the entire process from testing to deployment.

Blockworks co-founder Jason Yanowitz shares the judgment of Apollo's chief economist: agents may trigger bank runs by aggregating household cash.The chief economist stated that agents could transfer household cash from accounts with a national average interest rate of 0.1% to accounts paying 3% to 5%, causing banks to lose a significant portion of their cheap deposits.

Tom Lee, Chairman of Bitmine, stated that artificial intelligence and the cryptocurrency industry are gradually merging, with tokenization and AI agents likely becoming the two main driving forces of the next cryptocurrency market cycle.Tom Lee mentioned that BlackRock expects more assets to be tokenized, and financial settlement infrastructure is increasingly being built on cryptocurrency networks. Some fintech companies, such as Revolut, have already begun to build related businesses based on Ethereum. He also stated that AI is driving the formation of the digital economy, while cryptocurrencies provide the digital currency and blockchain infrastructure needed to support this economic system.

The Google Product Security Team has disclosed an internal AI agent called PageBreak, used to test the security of its first-party web applications. This agent is built on Google's Gemini model and began a pilot program in November 2025, transitioning to a formal project in January 2026, with the goal of autonomously scaling vulnerability discovery and reducing manual input.Unlike common AI scanning tools, PageBreak hands over hypotheses to specialized validators after discovering suspicious defects, attempting actual exploitation in a real-time running copy of the application, and only reports once confirmed exploitable, with a false positive rate close to zero. Google claims that PageBreak has identified over 500 XSS vulnerabilities in its first-party web applications, which can be used to hijack login sessions, steal data, or impersonate users.Google stated that the security team has been overwhelmed in recent years by a large number of AI-generated vulnerability reports that appear reasonable but are not valid, making it a major challenge to distinguish real defects from hallucinations. When testing applications built using the next-generation high-assurance framework, PageBreak found only two vulnerabilities. The next step for Google is to integrate PageBreak with the automated remediation agent CodeMender, providing confirmed vulnerabilities with accompanying fixes.