Building the protocol no single vendor can credibly own.
When an AI agent acts inside an enterprise system, the system needs to know, verifiably, offline, without trusting the agent's vendor, that a real human authorized the action. No platform vendor can be that verifier. The category needs a neutral protocol. We're building it.
Why Manav exists
In May 2026, the average enterprise runs 250,000+ non-human identities against perhaps 2,500 humans. By 2028, every Fortune 500 will deploy hundreds of autonomous agents. The compliance question every CISO and General Counsel is being asked, "who authorized this when the agent did the wrong thing?", has no good answer today. Session tokens don't prove human presence. Vendor-specific attestations don't verify across organisational boundaries. We're building the missing primitive: a fresh human signature at every consequential moment, with an offline-verifiable receipt any downstream system can check.
What "neutral" actually means
Microsoft can't credibly verify AWS agents. AWS can't credibly verify Anthropic agents. Anthropic can't credibly verify Salesforce agents. Verification by a competitor is verification you cannot trust. Every cross-vendor identity category in history has been won by a structurally neutral party, Visa for card networks, ICANN for DNS, Let's Encrypt for TLS, SWIFT for interbank messaging. Manav is being built to occupy that position for agent identity, with foundation governance, public IPR commitments, and the spec filed openly at the OpenID Foundation and IETF.
Open by default
The Manav Receipt format, the Delegation Chain extension, the verifier SDK, the conformance suite, and the reference implementation are all open source under Apache 2.0. The standards are filed at the OpenID Foundation and IETF with public IPR commitments. The commercial entity (Manav, Inc.) builds the issuer infrastructure, the operational dashboard, and the enterprise services. The protocol is the public good; the infrastructure is the business.
Get in touch
If you're a CHRO worried about deepfake hiring, a CISO worried about agent authority chains, a developer who wants to ship verified human consent in three lines of JavaScript, or a partner who'd like to embed Manav verification, email [email protected] and we'll respond within 24 hours.
For press, investor, or research enquiries: same address, prefix the subject with [PRESS], [INVESTOR], or [RESEARCH].