TL;DR
Meta, Walmart, Stripe, and Sierra have launched an open personal agent protocol to let AI agents authenticate with businesses safely, arriving just as Meta's Muse agent dominates app stores and Amazon's blockades expose the chaos of an unregulated agent web.
What happened
- Meta, Walmart, Stripe, and Sierra published a joint "personal agent protocol" on October 6, 2026, an open standard governing how AI agents interact with businesses.
- Bret Taylor, Sierra co-founder, OpenAI chairman, and former Salesforce co-CEO, is leading the initiative and spoke to CNBC about the effort.
- Meta's Muse, launched in early September, sits atop Apple's App Store above ChatGPT and already counts millions of U.S. users, per Meta VP David Singleton.
- The standard targets authentication and visibility: businesses can confirm an agent acts on behalf of a real person, and can see what that agent is doing on their platforms.
- OpenAI and Anthropic are not yet signatories, though Taylor said he expects them to join and would be "really disappointed" if competitors skip it.
Why it matters
- Amazon has already blocked both Muse and Perplexity's agents over scraping concerns, and sued Perplexity in November alleging concealment of bot activity, illustrating the legal and commercial risk of the current free-for-all.
- Taylor called the pre-standard environment "kind of chaos," and the protocol is designed to prevent unauthorized bots from masquerading as personal agents to access gated services.
- The OAuth analogy is deliberate: Taylor built Facebook Login at Facebook, and the group is pitching this as the same kind of universal credential layer, but for agents transacting on users' behalf.
- Singleton flagged that agents handling tasks like purchases and appointments require sharing credit card and personal data, making a security and visibility layer commercially necessary, not optional.
- If adopted broadly, the protocol shifts competitive leverage toward platforms that move fastest to implement it, and could freeze out agents (or companies) that refuse to comply with authentication norms.
What to watch next
- Whether OpenAI and Anthropic sign on: their participation would confirm this as a true industry standard rather than a Meta-led walled garden with friendly branding.
- Amazon's response: the company is the most visible holdout and has active litigation against Perplexity; its stance toward the protocol will signal whether the standard can bridge the retailer-agent divide.
- Adoption velocity among mid-market businesses: Walmart and Stripe provide retail and payments credibility, but broad SMB uptake is the real proof-of-concept for whether the "email for agents" analogy holds.
Originally published on Present of AI, a daily source-linked AI news timeline. Read the full timeline or browse the open dataset.