Agentic Identity
Identity for the age of AI agents.
As autonomous agents take actions on behalf of people and businesses, the web needs a way to know who's behind them. Dentity provides verifiable identity and scoped authority for agents, anchored to real, verified humans.
Principal
Alex Rivera
✓Verified human
Agent
research-agent.v1
✓did:dentity:agent:9f2c…
Sub-agent
summarizer.sub
attenuated scope
Authority flows down, and narrows at each step. Every link is cryptographically provable and instantly revocable.
Agent credentials
Issue cryptographic credentials to each agent describing who they are, who they act for, and what they can do.
Delegated authority
Let verified humans delegate scoped, revocable authority to agents, with expiration, audit trail, and granular permissions.
Provenance & trust
Every agent action carries a verifiable chain back to a real, verified principal. Trust, but verify, automatically.
Why it matters
When agents act, the web needs to know who's behind them.
AI agents are starting to browse, buy, sign, and transact on our behalf. But today there's no reliable way to answer the basic questions: who is this agent, who authorized it, and what is it actually allowed to do?
Dentity answers those questions cryptographically. Every agent is bound to a durable, verifiable identity and carries provable, scoped authority delegated by a real, verified human, or another agent.
Verifiable trust, end to end.
Dentity gives every agent a complete trust profile: not just an identity, but provable authority, scope, and accountability.
Agent Identity
Who is this agent?
Reputation
Can it be trusted?
Proof of Delegation
Who authorized it?
Scope of Action
What can it do?
Authorization Chain
Full provenance trail
Attenuation
Reduced privileges downstream
Auditability
Non-repudiation of actions
Revocation
Instant access removal
Cryptographic policies
Authority you can prove, and revoke.
We assign cryptographic policies to every agent call. Each policy is bound to the agent's durable identity, so the agent can prove it was delegated by a human or another agent. Like verifiable credentials, these policies are revocable, carry expiry, and are tamper-resistant. Only the controller of the private keys can present them.
Delegate
A verified human grants an agent scoped authority with defined limits and an expiry.
Act & prove
The agent presents its cryptographic policy on each call, proving who authorized it and what it's permitted to do.
Audit & revoke
Every action is verifiable and non-repudiable. Revoke authority instantly at any time.
Build trust into your agents.
Give your agents verifiable identity and provable, revocable authority, anchored to real, verified humans.