1Password for Agents：智能体用数字密钥自证身份、你在自己的设备上批准、1Password 代填登录——智能体全程看不到密码；Muse、Claude Duet、ChatGPT Work、dots、Grok Bot 与 Town 点名合作，而 Claude Duet 的登录本身仍带着 Anthropic 自己的「soon（很快）」

“1Password 宣布了一个让 AI 智能体登录你的账号、却永远看不到密码的办法——智能体用数字密钥自证、你在自己的设备上批准、由 1Password 代填。”

可用说法

• 2026 年 10 月 22 日（星期四；press 页面自己的日期标注「Thursday, October 22, 2026 9:00am EDT」；已按日历核算），1Password 宣布 1Password for Agents。按新闻稿：「1Password, a leader in identity security, today announced 1Password for Agents, a new platform that helps personal AI agents safely use the passwords, secrets, and credentials they need to complete approved tasks without seeing or storing them.（1Password——身份安全领域的领先者——今日宣布推出 1Password for Agents：一个新平台，帮助个人 AI 智能体安全使用完成任务所需的密码、机密与凭证，而看不到、也不存储它们。）」机制，按 press 页：「1Password for Agents lets AI agents sign in without receiving or seeing the credential. When an agent reaches a sign-in form, it uses a unique digital key to verify its identity to 1Password. The person sees which credential it needs and why, and then chooses whether to approve the request. That approval happens on the person's own device, confirming that the permission came from them. 1Password then fills it directly into the sign-in form instead of sending it to the agent. Once signed in, the agent continues the task while the password stays protected in 1Password.（1Password for Agents 让 AI 智能体在不接收、也看不到凭证的情况下完成登录。当智能体遇到登录表单，它用一个唯一的数字密钥向 1Password 验证自己的身份。用户会看到它需要哪个凭证、为什么需要，然后选择是否批准该请求。批准发生在用户自己的设备上，以确认权限确实来自本人。随后 1Password 把凭证直接填进登录表单，而不是发给智能体。登录之后，智能体继续任务，密码则始终保存在 1Password 里。）」它点名的问题：「Handing over a password, secret, or credential is a common shortcut to an agent, but can leave people with sensitive information in places they can no longer control as a copy gets saved for the next time.（把密码、机密或凭证直接交给智能体是常见捷径，但副本会被存下来，下次还在——你的敏感信息从此留在你无法控制的地方。）」生态：「1Password is designed to support people's AI agents of choice, and is building with companies across the personal AI agent ecosystem, including Muse, from Meta, Anthropic's Claude, OpenAI, SpaceXSI Grok Bot, and Town Townie.（1Password 的设计是支持用户自选的 AI 智能体，并正与个人 AI 智能体生态的公司合作构建，包括 Meta 的 Muse、Anthropic 的 Claude、OpenAI、SpaceXSI Grok Bot 与 Town Townie。）」合作方介绍包括：「Claude Duet works inside the tools teams already use to proactively get work done on their behalf. The 1Password integration gives Claude approved access when needed, so users can accomplish tasks without handing over their passwords.（Claude Duet 在团队已有的工具内、代表他们主动完成工作。1Password 集成在需要时给 Claude 经批准的访问，用户无需交出密码即可完成任务。）」；「ChatGPT Work and dots work with 1Password to sign into sites across the internet to help users get things done. Users approve sign-ins, and the passwords aren't exposed to the AI model.（ChatGPT Work 与 dots 与 1Password 配合，登录互联网上的各类网站、帮用户把事办成。用户批准登录，密码不会暴露给 AI 模型。）」；以及「Additional partners working on 1Password for Agents include Browserbase, Browser Use, Dana, Instinct, KERNEL, Perplexity, Ramp, Tomo, and Wajo, with additional integrations to come.（正在为 1Password for Agents 开发集成的其他合作方还包括 Browserbase、Browser Use、Dana、Instinct、KERNEL、Perplexity、Ramp、Tomo 与 Wajo，后续还有更多集成。）」Anthropic CPO Ami Vora 的引语：「Claude Duet asks before it signs in anywhere, and soon 1Password will complete the sign-in with the passwords people have already set up, so they don't have to start over.（Claude Duet 在任何地方登录前都会先问你；很快，1Password 将用人们已经设置好的密码完成登录，让他们不必从头再来。）」——这是页面上唯一的时间词，且是对 Duet 集成的未来时表述。1Password CPO Aashish Tripathi：「An agent should be able to do the job it's been asked to do without getting permanent access to everything behind the account, and 1Password provides the trust layer that makes this possible across personal agent experiences.（智能体应该能完成被交代的任务，而不必拿到账号背后一切的永久访问权；1Password 提供的信任层让这在各种个人智能体体验中成为可能。）」按 TechCrunch（Lucas Ropek，10 月 22 日下午 12:07 PDT）：「Password manager 1Password on Thursday launched a new tool by which users can allow agents to log into their personal accounts on their behalf, without the agent ever having access to a user's credentials.（密码管理器 1Password 周四发布了一个新工具：用户可以让智能体代表自己登录个人账户，而智能体永远接触不到用户的凭证。）」；以及该媒体的机制转写：「When an agent attempts to log in on a user's behalf, a unique digital key is sent to 1Password to confirm the agent's digital identity. The user is then sent an alert about the login attempt and can personally confirm the request. 1Password then signs in the agent, filling in the credentials on the digital assistant's behalf, which allows it to log in and complete whatever task is at hand.（当智能体试图代表用户登录时，一个唯一的数字密钥被发送给 1Password 以确认智能体的数字身份。用户随后收到关于这次登录尝试的提醒，可以亲自确认该请求。之后由 1Password 替智能体登录、代为填入凭证，让它能登录并完成手头的任务。）」TechCrunch 的同日背景：「Indeed, 1Password's new platform is launching on the same day that Axios reports that U.S. banks have begun to detect agentic activity inside user accounts.（1Password 新平台发布的同一天，Axios 报道称美国银行已开始检测用户账户内的智能体活动。）」——银行层内容经由 TechCrunch 对 Axios 报道的转述；Axios 本 run 未被打开。