Central Identity Database
One SignInOnce sign-in, one master persona, and AI agent personas that act only within the permissions you give them. This page keeps what CID does today apart from where we're going.
Abstract
The Central Identity Database (CID Global) is the identity layer behind SignInOnce. It holds who you are, the AI agents you create, and what each of those agents is allowed to do.
This whitepaper is in two parts. What CID does today describes only what runs now. Where we're going describes what we are building next; none of it is available yet.
- Today: one sign-in and one master persona, with AI agent personas that sit under it
- Today: identity checks (KYC levels 0–3) reviewed by our staff
- Today: agents act only through tools you allow, with a human approval by default
What CID does today
Everything in this part runs now.
Identity and sign-in
One account, one master persona
You sign in once with SignInOnce, and the apps that connect to SignInOnce recognise you. CID gives every account one master persona, which is you, and keeps the AI agent personas you create under it.
Single sign-on
One SignInOnce login for the apps connected to it, using the OpenID Connect standard
Not tied to your email
Your account is identified by its permanent SignInOnce ID, not by an email address
Master persona
One master persona per account; your AI agent personas sit under it
KYC levels 0–3
Level 1: a verified phone, your name and an address on file. Level 2: adds an ID document. Level 3: adds proof of address and source of funds. Documents are reviewed by our staff, and a second, different reviewer can be required.
AI agent personas
Agents with their own identity and only the permissions you give them
You can create AI agent personas under your master persona. Each agent has its own identity in CID, a name and role you choose, and a list of the tools it is allowed to use. An agent cannot use a tool that is not on its list.
How agents are kept in check today
- •Tool permissions: each agent has its own list of allowed tools
- •Human approval by default: unless you choose otherwise, an agent needs a person to approve each tool call
- •Trust per tool: an agent can earn automatic approval for one tool after a run of clean uses; a single failure puts it back on approval, and tools marked high-stakes always ask
- •Kill switch: one agent tool, or every tool covered by the global switch, can be stopped at once
- •Vault: secrets your agents use are kept in the CID vault, encrypted on our servers, and the console shows only masked details
Companies
You can group your agents under a company you own, marked as a real or a virtual company
Roles and capabilities
Each agent carries a role and capability tags you set, so you can see at a glance what it is for
Security and privacy today
Data protection
- • Encrypted connections (TLS) between your device and our servers
- • Agent secrets in the CID vault are encrypted at rest
- • Not zero-knowledge today: our servers can read the data you give us. How we plan to change that is under “Where we're going”
Access control
- • Each connected app gets its own key for KYC data, limited to what that app needs: one may report a verified phone, another an address on file, another may only read your KYC level
- • Per-agent tool permissions, with human approval by default
- • A kill switch for agent tools
Where we're going
What we are building next. Nothing in this part is available today.
Your documents, your key
- • ID documents encrypted on your own device, with a key only you hold
- • A signed KYC result that you carry and can show to an app, so the app does not need your documents
- • The copy the law requires us to keep stays sealed, and opens only when two of three named key holders act on a supervisor's request
Easier verification and contact
- • Verify your identity with an eID (itsme)
- • Contact us through a NextICS link
- • A complete audit log of every agent action
Agents that do the work
- • Agents that answer messages, plan meetings and compile research for you
- • Agents that work with each other and with people
- • Ready-made agent roles, such as research assistant or customer support
Organisations of agents
- • Agent teams with a structure: a lead agent, team agents and specialists
- • Permissions set per company, not only per agent
- • Manage your accounts and your agents from one place
Example scenarios
Illustrations of where we're going, not services available today
Solo Entrepreneur
- • Sales Agent: answers inquiries
- • Admin Agent: prepares invoices for the owner to approve
- • Research Agent: keeps track of the industry
Small Business
- • Customer Support: first-line answers, handing over to a person when needed
- • Data Analysis: regular reports
- • HR Assistant: help with onboarding
Property Owner
- • Tenant Relations: handles questions and maintenance requests
- • Market Analyst: tracks the market
Personal Life
- • Personal Assistant: schedules and reminders
- • Travel Planner: research for your trips
The direction
Where we want CID to go:
Portable identity
One identity you carry across the services that connect to CID Global
AI that works for you
Agents that take on real work, inside limits you set
Privacy by design
Your documents encrypted with your key, and apps that need only the result of a check
In development
CID Global is in development. This page will be updated as roadmap items become available.
Contact via NextICS (coming)