Meta and Sierra unveil the Personal Agent Protocol: an open standard for how personal AI agents interact with businesses
In Sierra's own words: Meta and Sierra are developing the Personal Agent Protocol with industry partners at Genesys, Instinct, Rocket, Shopify, Stripe and Walmart, an open standard that defines how personal AI agents interact with businesses. It is designed to handle authentication, put authorization in consumers' hands, and give companies visibility into what personal agents do through their websites, APIs or company agents. Sierra states plainly that it is open for anyone to implement.
The problem is stated plainly by the people building it. Per Sierra's official blog, most personal agents today use websites and apps the way people do — loading pages and clicking through forms. When that cannot get the job done they may call the company's support line or open its web chat, which the post says takes a long time and may still fail, whereas a direct connection could get the same task done securely in seconds. For that connection to be adopted at scale it has to work for all three parties: consumers want speed, dependability and trust, for the job to be done right the first time; brands want visibility and control, to know when a personal agent is acting for a customer and to decide for themselves what it can do; and companies building personal agents want efficiency and access, a direct, consistent way to work with participating companies.
The principle in one line: consumers grant access, companies set the boundaries. Per Sierra's blog, the principle behind the protocol is that consumers decide what access to give their personal agents, and companies set parameters for what those agents can do. The flow works like this: the protocol starts on the website, where a personal agent can discover what the company offers and how to reach it, then begins a session on its user's behalf. That session can start as a guest, which may be enough to check product availability or ask about a returns policy. When a task requires access to a customer's account, the customer can sign in on the company's page or use credentials already set up with their personal agent. Sierra stresses the customer is always in control, deciding whether the agent has read-only or write access.
What it is built on technically. Per Sierra's blog, the session is built on OAuth, an established standard for authorizing access. It carries across channels, so a question asked before sign-in and an order change made afterward are part of the same visit rather than two unrelated operations. On top of that, a personal agent can get the job done using whichever route the company believes offers the best customer experience; Sierra names three: the company's website, by navigating its regular web pages; its APIs, by connecting through interfaces built on standards such as MCP and OpenAPI; or its own agent, for tasks that need conversation such as a warranty claim. Sierra's summary: the company decides what it makes available, the personal agent gets a consistent way to connect, and the customer gets a faster way to get things done.
Next steps and the longer road. Per Sierra's blog, the company wants to develop the protocol with the companies and personal agent builders using it, and announced that Instinct will also be joining the effort. It welcomes all partners and plans to publish the v0.1 specification later in October 2026, host design workshops with interested parties, and publish a reference implementation to help developers get started. Sierra also sets out what comes later: more detailed permissions that let customers and companies set limits on specific actions; push notifications that let a company tell a personal agent the moment a flight is delayed or an order ships; and payments extensions that let a personal agent complete a purchase without sharing credit card information. All three point the same way: turning whether an agent may do a given thing on my behalf from a vague customer-service promise into a configurable, auditable interface contract.
Who is pushing it, in their own words. Per Sierra's blog: Genesys Chairman and CEO Tony Bates says personal AI is creating a new front door to the enterprise, and brands need a trusted way to know who an AI agent represents, what it is authorized to do, its intent, and how to work with it securely; Rocket Companies CTO Shawn Malhotra says Rocket and Redfin are building for a world of persistent agents taking action, and agentic technology developed with Sierra lets a Muse agent move across their platform from finding a home to securing financing; Shopify VP Product Mani Fazeli says that as personal agents become part of everyday life merchants have a new frontier for customer service, from answering product questions to supporting purchases, returns and exchanges; and Stripe's Head of Payments Kevin Miller says a customer relationship does not start or end at checkout, and when customers send an agent they expect the same service they would get themselves. Per The Verge's October 6, 2026 report, Genesys, Instinct, Rocket, Shopify, Stripe and Walmart are also helping develop the protocol.
🤔 Frequently Asked Questions
What is the Personal Agent Protocol?
Per Sierra's blog, PAP is an open standard Meta and Sierra are developing with industry partners that defines how personal agents interact with businesses: it handles authentication, lets consumers decide how much access their agents get, and gives companies visibility through their websites, APIs or company agents. Sierra says it is open for anyone to implement.
Who is involved?
Per Sierra's blog, beyond Meta and Sierra the industry partners are Genesys, Instinct, Rocket, Shopify, Stripe and Walmart, with Instinct additionally named as newly joining the effort.
What is it built on, and how does an agent complete a task?
Per Sierra's blog the session is built on OAuth and carries across channels; tasks can be completed through three routes: the company's website, its APIs built on standards such as MCP and OpenAPI, or the company's own agent for conversational tasks such as a warranty claim.
When can developers use it?
Per Sierra's blog, the plan is to publish the v0.1 specification later in October 2026, host design workshops with interested parties and publish a reference implementation to help developers get started.
What capabilities come later?
Sierra names three: more detailed permissions so customers and companies can limit specific actions, push notifications so a company can alert an agent when a flight is delayed or an order ships, and payments extensions that complete a purchase without sharing credit card information.
🛠️ Recommended Tools
- AI Agent Interoperability Tool Interoperability is the whole point of PAP: one agent has to reach countless companies' websites, APIs or company agents. Sort out interface alignment before talking about access.
- AI Agent Sandbox Sierra leaves read-only versus write access to the user and plans finer permissions later. Verifying that at design time needs an environment where the agent can be boxed in first.
- AI LLM Router When one session continues across websites, APIs and company agents, the underlying model calls get scattered. Routing decides which model handles which step and where the cost lands.
Summary
Per Sierra's official blog post Introducing Personal Agent Protocol (October 6, 2026): Meta and Sierra are developing the Personal Agent Protocol (PAP) with industry partners at Genesys, Instinct, Rocket, Shopify, Stripe and Walmart, an open standard defining how personal AI agents interact with businesses, open for anyone to implement, designed to handle authentication, empower consumers and give companies visibility through websites, APIs or company agents. Sierra states the approach: consumers decide what access to give their personal agents, companies set parameters for what those agents can do; the session starts on the company's website, can begin as a guest, and when an account is needed the user signs in on the company's page or uses credentials already set up with the agent, deciding whether access is read-only or write; the session is built on OAuth and carries across channels; and tasks can be completed through the company's website, its APIs built on standards such as MCP and OpenAPI, or the company's own agent. Sierra plans to publish the v0.1 specification later in October 2026, host design workshops and publish a reference implementation, with later directions including more detailed permissions, push notifications and payments extensions. Per The Verge's October 6, 2026 report, Genesys, Instinct, Rocket, Shopify, Stripe and Walmart are also helping develop it. Every fact here comes from the official source listed below plus the media coverage named inline, with no speculation added.
Sources: "Sierra(官方): Introducing Personal Agent Protocol"
"The Verge(2026-10-06): Meta is helping develop an AI agent protocol"