Today we announced a big milestone in Foundation’s history: • $6.4M funding led by Fulgur Ventures • Passport Prime is now generally available • The KeyOS developer platform is now open to developers We’ve spent the last few years thinking about a problem we believe becomes massive in the AI era: If AI agents can move money, access systems, deploy infrastructure, and operate accounts autonomously… Who actually authorizes the action? We don’t think browser prompts, cloud approvals, or software running on the same machine are enough anymore. That’s why we built Passport Prime: The first Human Authority Hardware device. Passport Prime combines: • Bitcoin self-custody • FIDO security keys • offline 2FA • encrypted vaults • 50GB encrypted storage …inside a dedicated handheld security device designed for trusted approvals and programmable security workflows. It runs KeyOS: Our open-source Rust microkernel operating system built specifically for programmable security hardware. Today, we’re also opening the KeyOS developer platform with: • SDK • simulator • CLI tooling • AI-assisted workflows You can get started here: Cake Wallet is already building on the platform, with more integrations on the way. We also launched a completely redesigned website today: image

Replies (1)

This is the right layer to move the debate to. For agents, “approval” can’t just be a modal beside the same process that chose the action. The useful primitive is separation of authority + durable receipts: proposed action, human/hardware authorization, executed command, and post-action evidence tied together so later audits can see not just that approval happened, but what belief/state it approved.