agent

According to official news, the on-chain smart money signals of the Binance wallet now support automated trading through Agentic Wallet. Users can utilize the "On-chain Copy Trading Skill" to allow the Binance Agentic Wallet to automatically buy when corresponding signals are detected, and simultaneously set stop-loss and take-profit rules.The operations listed by the official source include: configuring and installing the on-chain copy trading Skill for the Agent; selecting the on-chain smart money signals of the Binance wallet as the trading trigger source; setting the target blockchain, trading budget, position size, and take-profit and stop-loss rules; first conducting a free simulation run and observing the strategy's performance, adjusting parameters based on the results, and confirming before switching to live trading.The official source stated that the entire trading process can be fully automated and run around the clock. Once users set a trading strategy, subsequent copy trading will be executed automatically by the system.

According to the Coinbase blog, its MCP service Coinbase for Agents has launched Agent Controls, allowing users to set limits, product permissions, and stop conditions for trades conducted through the AI Agent. The relevant control settings run on Coinbase's server side, and users must log into their accounts via the Coinbase App or web to manage them; the AI Agent cannot modify or bypass them on its own.The features launched include: Maximum trade size, which sets the maximum amount for each trade in USD; 24-hour volume limits, where trades exceeding the limit cannot be executed; Auto-stop, which will stop the AI Agent from placing orders when the portfolio loss exceeds a set percentage, until the user re-enables trading in the App.In addition, the product permissions feature will soon be available, allowing users to choose the categories of products that the AI Agent can trade, including cryptocurrencies, stocks, and derivatives.

Stablecoin payment infrastructure provider Rain has launched the AI agent payment product Agent Vaults. Businesses can embed payment and receipt functions into applications using their own branding, allowing AI agents to collect payments, shop, and transfer funds on behalf of users or businesses within authorized limits.The product supports funding through bank transfers, payment cards, or stablecoins, with funds able to be stored as fiat currency in bank accounts or stablecoins in embedded wallets. Agents can make payments using Visa or Mastercard, as well as remit to bank accounts or directly transfer stablecoins. Users or platforms can set spending limits, designate merchants or payees, and restrict payment times.

The locally developed AI Agent Alma, led by independent developer Yetone, ranked third in the latest round of cross-evaluation initiated by the AI Agent infrastructure company Runta, both in task success rate and cost of success, outperforming tools like Claude Code and OpenCode. The current leaderboard snapshot shows that Alma has a task success rate of 72.7%, just behind Codex CLI at 77.9% and Pi at 72.8%, while Claude Code stands at 70.0%.Alma costs approximately $0.80 on average for each successfully completed task, compared to Claude Code's $0.94, making it about 16% cheaper. This FrontierHarness v1.1 comparison evaluates the performance of the same large model under different Agent execution frameworks, covering 21 sets of Harness configurations, 5 large models, and 62 tasks, with each task repeated 4 times, planning to run a total of 26,040 times, which is 72 times the scale of the previous version. All tools operate in a unified cloud environment, comparing task success rates, token costs, and execution speeds.Alma emphasizes local operation and long-term memory, capable of reading and modifying code, executing terminal commands, searching the web, and supporting Skills and MCP, while also allowing for switching underlying models. Recently, Alma integrated with Herdr, allowing programming Agents like Claude Code, Codex, and Pi to continue operating within Herdr, and enabling users to view execution processes, respond to questions, and approve operations in a centralized chat interface. Users can also remotely manage multiple Agents on their computers via iPhone. The evaluation is still ongoing, and rankings may change; the models being tested are limited to open-source models like GLM, DeepSeek, Kimi, and Muse, and do not cover closed-source models from the GPT and Claude series.

Google's AI programming tool Antigravity has added a Knowledge Work office mode, making the AI Agent, originally aimed at programmers, easier to use for research, writing, data analysis, and creating documents and presentations.The new mode hides development tools such as terminals, Git, and code modification records, focusing instead on conversations, documents, and task outcomes. Users do not need to understand code repositories and can directly have the Agent handle local files, call tools, and complete multi-step tasks.The old and new modes use the exact same underlying Agent, and there has been no reduction in models, plugins, Skills, MCP, or custom Agents. Users can switch between modes at any time without needing to restart the application. Google revealed that currently, tens of thousands of non-software engineering employees within the company use Antigravity weekly, covering departments such as sales, finance, law, design, research, and operations.

On November 10, the B.AI platform announced the official launch of Anthropic Claude Fable 5.2, with API and Web Chat available simultaneously. This model is designed for complex reasoning, software engineering, and long-cycle agent workflows, always enabling adaptive thinking, equipped with a 1M Token ultra-large context window and a maximum output of 128K Tokens.According to the Artificial Analysis Intelligence Index v5 rating, Claude Fable 5.2 topped the list with a score of 65 under the highest configuration, far exceeding the median score of 40 among similar models. The generation speed has improved by up to 30% compared to Fable 5.1, with a standard price of $8.00 per million input Tokens and $40.00 per million output Tokens, and it supports prompt caching. You are welcome to log in to chat.b.ai/chat to experience it.

According to a report by Decrypt, the blockchain data project Space and Time, supported by Microsoft, announced the launch of the Space and Time Enterprise platform, which allows enterprises to use private data for AI tools without relinquishing the underlying records. The platform can prove that computations based on private data are correct, enabling AI applications, auditors, and other relevant parties to verify results without accessing the database itself.The first batch of partners includes on-chain asset management firm Maple Finance, Indonesian automotive manufacturer and distributor Indomobil, Oracle software consulting company Trenez, and The Data Foundation, a nonprofit organization focused on government data based in Washington, D.C.Nate Holiday, CEO of the original developer of Space and Time, MakeInfinite Labs, stated that enterprises want AI agents to utilize this information to execute workflows, but the agents do not need unrestricted access to the records behind every computation.He mentioned that granting unrestricted data access to AI tools would create new exposure channels. The company is collaborating with clients in construction, engineering, and manufacturing to enable agents, reporting tools, and compliance teams to work based on the same set of business records and verify the computations they rely on.The platform is based on Space and Time's patented technology, Proof of SQL, which is used to prove that database queries are computed correctly and is currently compatible with Google Cloud's BigQuery. Enterprises can run it on their own servers without needing to operate a blockchain, purchase dedicated hardware, or rely on external proof services.With GPU acceleration, supported queries on millions of rows of data can be completed in one second and generate proofs. This technology can only prove that computations are correctly executed on given records and cannot determine whether the records themselves are accurate. Space and Time completed a $20 million strategic financing round led by Microsoft's venture fund M12 in 2022, and Grayscale previously launched a trust product for the SXT token.

According to Bloomberg reporter Mark Gurman, Apple's new CEO John Ternus recently spoke with software chief Craig Federighi, requesting to accelerate AI development and no longer be satisfied with merely patching Siri's existing features. Apple's next goal is to transform Siri into a truly proactive agent.Currently, Siri can understand screen content, read user-authorized emails and messages, and perform some operations across applications, but it essentially still listens to one command and executes one action. Apple hopes it can remember tasks assigned by users over the long term, like ChatGPT Dots and Meta Muse, proactively check for changes, determine what to do next, and continue to push forward in the background.Devices like the iPhone, Mac, and Apple Watch carry a large amount of personal information, and Apple also controls the operating system and application calling interfaces. Earlier this year, Apple announced that the number of active devices worldwide has exceeded 2.5 billion, but not all devices support the new version of Siri. Gurman stated that Apple could launch a version closer to a personal agent as early as 2027; if issues regarding autonomous planning, background execution, and permission security are resolved, Siri has the opportunity to become the personal agent entry point on iPhone and Mac.

Phoenix Technology reported on November 9, Beijing time, citing MacRumors, that compared to the traditional version of Siri, Apple's revamped Siri AI has significantly improved in terms of intelligence and personalization capabilities, but it still falls short of proactive, always-on agents like OpenAI's Dots and Meta's Muse. Bloomberg reporter Mark Gurman stated on Sunday that he learned Apple may enable Siri's features to resemble an intelligent agent as early as next year.Gurman mentioned that Dots, Muse, and Grok can continuously track targets, determine what to do next, and continue to advance established tasks without users needing to issue commands one by one. He stated that he was informed Apple ultimately hopes to develop in this direction and believes that Apple may enable Siri to become more like an intelligent agent as early as next year. Gurman also noted that in a closed-door meeting in recent weeks, Apple CEO John Ternus told software engineering chief Craig Federighi that Apple needs to accelerate the pace of its software roadmap and quickly move beyond merely making necessary improvements to Siri.

According to reports from multiple media outlets including Decrypt, the U.S. Attorney's Office for the Eastern District of Virginia stated that former FBI agent Patrick Steven Yaroch pleaded guilty on November 5 in federal court in Alexandria to charges of telecommunications fraud and money laundering.He used stolen credentials to conduct approximately 100 fraudulent transfers from 35 cryptocurrency wallets involved in FBI investigations, stealing about $2.47 million in cryptocurrency assets, and concealed the source of the funds through methods such as exchanging tokens. Some of the funds were used to pay off debts such as credit cards and auto loans. Yaroch faces a maximum of 20 years in prison, with sentencing scheduled for March 4, 2027. Previous reports indicated that the amount involved was approximately $1 million.

AI search company Perplexity officially launched Teammates, allowing AI Agents to be responsible for specific tasks long-term like employees. CEO Aravind Srinivas revealed that approximately 10,000 AI Teammates have been created internally in the last 30 days, which is 20 times the total number of human employees.Teammates are built on Perplexity Computer. After users assign roles, goals, and permissions to the AI, it can continuously track work progress, proactively identify issues, and call upon tools to address them. Multiple Teammates can also divide tasks among themselves and contact human colleagues via Slack. For example, a sales Agent can discover that a customer has not placed an order for a long time, automatically investigate the reasons, and prepare a follow-up plan; a development Agent can hand off the task of writing a fix code to another Agent after discovering a website bug. Each Agent has an independent working environment, retaining project memory and task history, allowing users to check progress, adjust direction, or take over operations at any time.Business Insider exposed Perplexity's internal Teammate tool in July this year, which was primarily used for software engineering tasks such as bug tracking and system monitoring. The product has now expanded to marketing, sales, customer service, and operations, adding long-term task management and cross-Agent collaboration. Teammates are currently available in research preview form to Max subscription users at $200 per month, initially supporting the web version, with native mobile and desktop versions to be launched later.

According to Cointelegraph, VeChain co-founder and CEO Sunny Lu recently stated in Cointelegraph's Chain Reaction program that autonomous AI agents operating around the clock require a "new accountability system" to establish public trust, and that "the perfect solution is blockchain."He made this statement while introducing the AI agent platform VeWorld.ai built on the VeChain blockchain. According to his vision, AI agents will have digital "passports" and operate within a "decentralized reputation system" that tracks the reputation of agents and their creators, while smart contracts can help settle transactions between agents or between agents and people. VeChain has previously advanced blockchain applications through supply chain tracking, including food traceability projects.Lu believes that AI agents can also help individuals convert their expertise into businesses or services, expanding software development opportunities to non-developer groups. He likened this opportunity to mobile app stores, which allow individual developers to distribute and monetize their creations, while agent-building tools can open this market to people beyond programming skills, including those with professional experience, local knowledge, or hobbies.He himself has undertaken several initiatives, including creating an educational agent for his daughter, an agent to support his mother's daily routine and rehabilitation, and a virtual employee to manage procurement, inventory, sales, and operations for a shop owned by an acquaintance in Milan.Meanwhile, other industry participants have pointed out the risks of relying on agent service providers. Animoca Brands Chairman Yat Siu stated in an October interview with Cointelegraph that while crypto traders can have AI agents work around the clock, he warned users that they might build trading strategies around services they cannot control.

On November 5, B.AI announced that it now supports account top-ups via API. Developers can directly integrate the top-up capability into their self-built Agents and automated workflows, achieving automated replenishment of computing power and quotas, reducing manual operations and interruption risks, and maintaining stable connections for high-frequency calls, batch tasks, and long-running Agent services.This capability further completes B.AI's automation ecosystem for developers, allowing model calls, cost management, and quota replenishment to be accomplished through API, making it easier for teams to embed AI capabilities into their business processes. Welcome to visit b.ai for more information.

According to TNW, the Barcelona startup Kabilio has completed a €10 million financing round, led by Picus Capital and Visionaries, with participation from the Spanish public agency for technology and innovation CDTI.The two leading investors also participated in the company's €4 million pre-seed financing round a year ago. Kabilio aims to replace processes such as debt collection in accounting firms with AI agents.

The Dark Side of the Moon announced the release of Kimi Hosted Agents, which will provide the Harness that supports Agent operation as a cloud service for enterprises and developers. Enterprises can obtain a 24/7 managed environment without the need to build and maintain their own operational framework, supporting long-range task execution and efficient collaboration among multiple agents.

Circle CEO posted on the X platform that Circle has launched a brand new Agentic Managed Payments, upgraded and expanded the agent market, while providing simpler tools for agents to obtain funding.