Evergreen Tools
|
← Back to AI Tools

Persona AI: Identity Review Agents You Can Audit

Persona launched Persona AI on 29 October 2026: Persona Assistant answers plain-language questions over identity data, Case Review Agents gather evidence and check it against your own policies before recommending a decision with visible reasoning, and Persona Extend connects through MCP and CLI connectors. Built and tested with more than 30 early-access customers including Zillow and Etsy

Official Links

官方博客 · Business Wire

Features

  • ✓ Case Review Agents gather evidence, check it against your own written policies, then recommend a decision with visible reasoning — clear cases auto-approve, exceptions go to a person
  • ✓ Persona Assistant answers plain-language questions about your identity data, such as why verification rates are trending up, and completes tasks from the dashboard with approval at the steps that need it
  • ✓ Policy Agents turn compliance and risk requirements into a full workflow, flagging gaps and conflicts with existing rules before anything reaches production
  • ✓ Persona Extend uses MCP and CLI connectors so teams keep using their preferred AI tools to build and configure identity workflows, with a traceable record of every action
  • ✓ Human oversight is designed in: every action is transparent, permissioned and auditable, and sensitive PII stays inside the platform boundary

How to Use

  1. Map your current review process and separate the parts that are pure evidence collection and policy comparison — those are the first candidates for an agent
  2. Write your review policy out as structured clauses and feed it to Policy Agents first, letting it surface conflicts and gaps in the existing rule set
  3. Run shadow mode on a low-risk queue such as routine KYC re-checks: the agent recommends, a human decides, and you compare where the two disagree
  4. Once disagreements converge, open up auto-approval thresholds while keeping audit trails and an exception escalation path so every decision traces back to evidence and a policy clause

FAQ

What is included in Persona AI?

Per the announcement, Persona AI is a suite of three: Persona Assistant, Persona Agents and Persona Extend. Assistant answers identity data questions in plain language and executes tasks with approvals; Agents includes Case Review Agents and Policy Agents for gathering evidence, checking policy and recommending decisions; Extend connects your existing AI tooling through MCP and CLI connectors. Details at https://withpersona.com/blog/introducing-early-access-for-case-review-agents

Do Case Review Agents make decisions on their own?

The design deliberately separates recommendation from decision. The agent gathers evidence, checks it against company policy and recommends an outcome with its reasoning visible; cases meeting the team criteria can be approved automatically while exceptions go to a human. Persona CEO Rick Song frames it as AI plus the judgement of identity experts, with people firmly in charge of the decisions that matter. In practice you define which decisions an agent may recommend, which require human approval, and what evidence must be retained

Is there any real production data?

Two published numbers are worth noting: Persona says Persona AI already powers hundreds of thousands of identity decisions in production, and Parafin cut median KYC review time for capital products by about 95% after Case Review Agents took over routine reviews, letting its risk team focus on fraud investigation. Persona also reports deepfakes on its platform growing more than 50x in recent years, with attacks outpacing hiring — the stated reason for pushing AI into identity operations. These are vendor figures; check the original announcement for exact definitions

Can a company without existing AI infrastructure use it?

Officially, no existing AI setup is required: Persona AI runs inside the Persona platform where identity data, policies and case history already live, so you do not build the identity context and PII controls yourself. If you already verify identities with Persona, adoption mostly means structuring your review policies and setting auto-approval thresholds. If you are not on Persona, the first question is whether you want to move identity review onto this platform at all

What compliance credentials back it?

Persona is an established identity platform and Persona AI inherits that foundation, with the official emphasis on transparent, permissioned, auditable actions. Be clear that regulatory responsibility in your sector and region — KYC/KYB obligations, for instance — remains yours. Before going live, align audit logging, data retention and the human-review ratio with legal. Official site at https://withpersona.com/