Skip to content
↑↓ to move · Enter to openEsc to close
Example verification flowScenario 07

AI recruiting agents: verifying the candidate's agent and the employer's

Mike asks his job-search agent to apply for an operations manager role and set up a first call. The employer runs a screening agent that handles applications at volume. Each side has a reason to ask who is on the other end: the employer does not know which agent sent the application, and Mike's agent does not know whether the "employer" is real.

This scenario shows verification running in both directions, which matters in hiring because impersonation happens on both sides.

Ready
Human
Mike
“Apply to the operations manager role and set up a first call.”
Candidate’s agent
Identity proof
Mike's job-search agent
Operator: Mike (individual)
“I’m applying on behalf of Mike.”
Employer screening agent
Brennan Outdoor Co.
Recruiting and screening agent
Who are you?
Example verification flow
Fresh challengenonce 4C15·8F
Proof of control✓ VALID
Operator declarationL1_REGISTEREDDECLARED
Employer’s domain claimL2_DOMAIN_VERIFIED✓ VERIFIED
StatusACTIVE
Checked separately
Qualifications and referencesEMPLOYER’S SCREENING

Identity is not permission, and it is not a credential. AgenID says which agent applied and who declared it; qualifications, references, eligibility and the decision to hire are assessed by the employer.

Awaiting proof
Nothing proceeds on an unproven claim
What AgenID proves here

Which agent applied and who declared it, and, in the other direction, that the employer’s agent controls the domain it names. Qualifications, references and the hiring decision are the employer’s to assess.

Illustration, not deployment

These scenarios illustrate the protocol's verification model. The reference deployment at agenid.com issues nothing above L1_REGISTERED today: L2 and higher require a VerificationAssertion signed by a root authority key, and that key has not been created. L5 is reserved by the spec and is not issuable at all.

Challenge–response is a proposed interaction pattern. v1.1.1 defines the signed objects, not this exchange.

Why hiring needs identity on both sides

Employers see automated applications they cannot attribute, and candidates are targeted by fake recruiters who collect personal details under a real company's name. A screening agent that cannot tell one applicant's agent from another, and a candidate's agent that cannot check the employer, both end up guessing.

Mutual verification before the screen

The candidate's agent presents its AgenID. The employer's screening agent resolves it and re-verifies the operator's signature; the operator declaration shows as DECLARED at L1_REGISTERED, which is what the candidate says about their own agent. In the other direction, Mike's agent resolves the employer's agent and sees a domain claim at L2_DOMAIN_VERIFIED, meaning an authority checked that the employer controls the domain it names.

Neither side depends on a job board or a shared platform to do this. Each resolves the other's public record on its own.

Identity is not a credential

A verified agent identity says nothing about a candidate's qualifications, references or eligibility to work, and nothing about whether an offer should be made. Those are assessed by the employer through its own screening. AgenID answers who is on the other end of the conversation.

Today the reference registry at agenid.com issues DECLARED and L1_REGISTERED. The higher levels shown in this illustration require an assertion signed by a root authority key that has not been created yet, so they describe how the protocol works rather than something you can obtain today.

Frequently asked questions

Can an AI agent apply for jobs for me?

Many can fill in applications and schedule calls. What the employer usually cannot tell is which agent is applying and who stands behind it, and what the candidate cannot tell is whether the employer's agent is genuine.

How can a candidate's agent check a recruiter is real?

By resolving the recruiter agent's AgenID, re-verifying its operator's signature, and checking whether a domain claim for the employer has been independently verified. Domain verification is part of the model; it is not issuable today.

Does a verified identity mean the candidate is qualified?

No. AgenID answers which agent is acting and who declared it. What that agent is allowed to do stays with the business on the other side and with the operator that runs the agent. Protocol v1.1.1 defines no signed authorization object.

Does AgenID store résumés or candidate data?

No. The registry holds the agent's signed manifest and public key. Candidate materials go wherever the candidate and employer send them, outside AgenID.

Integration patterns for this scenario

How the identity travels through platforms agents in this scenario commonly run on. Each brief is a pattern using the platform’s documented APIs, not a shipped adapter package.

Give your agent an identity

Register an agent in the browser — the signing key is generated on your device and never sent to AgenID — or resolve an existing AgenID and re-check its signature yourself.