# User lifecycle

Invitations, account states, the posture risk score, admin impersonation, and session management.

A user account is created by an invitation or a self-service signup, moves through a defined set of states, and carries a posture risk score that reflects its exposure. Administrators can act on a user's behalf and end their active sessions. All of these lifecycle actions are reached from a user's detail page in [People](https://dts-identity.com/docs/directories/people.md).

## Invitations

An invitation lets an external user create their own account instead of you creating it for them.

1. From the People list (or from an [organization](https://dts-identity.com/docs/directories/organizations.md)), choose to **invite** a user and enter their email address.
2. Pick the target directory (one that allows signup), and optionally an organization, organization-admin rights, and applications to pre-authorize.
3. Send. The recipient gets an email containing a one-time **invitation code**.
4. The invited user follows the link, completes their profile, and sets their credentials — the account is created on completion.

In a directory that has [passwords switched off](https://dts-identity.com/docs/directories/user-directories.md#turning-passwords-off-password-optional-directories), step 4 asks for no password. The invited user completes their profile and from then on signs in with an emailed code.

Pending invitations are listed on the **Invitations** page, where you can **resend** (a fresh email), **cancel**, or **delete** them. Invitations **expire after seven days**; expired invitations are removed automatically after a further retention period. Resending an invitation refreshes its validity.

## Self-service signup

When a directory allows signups, people can register themselves from the sign-in screen. By default such directories also **require email verification**: the account is not usable — and not even created — until the address is confirmed.

1. The person fills in the signup form. Their details are held temporarily; **no account exists yet**, and the typed address is not written anywhere on a user record.
2. A verification link is sent to that address. The link is valid for three days.
3. Opening the link shows a confirmation page. Choosing **Continue** there **creates the account** with the email already verified and the status set to **Active**, then the person can sign in. The extra click exists so that mail security scanners, which follow links in incoming mail before the recipient does, cannot use up the one-time link.

Until the link is opened, a sign-in attempt with that address simply fails — there is no account to sign in to. If the email never arrives, the **Resend verification email** link on the sign-in screen sends a fresh link; the response is the same whether or not a registration is pending, so it never reveals which addresses are in use. A pending registration that is never confirmed expires on its own after three days, leaving nothing behind.

This gate applies only to self-service signups. **Invitations** are exempt — the invited address is already trusted, so an invited user is created immediately with a verified email. Synchronized and administrator-created users are also unaffected.

You can turn the gate off per directory with **Require email verification** under the directory's self-service settings (see [User directories](https://dts-identity.com/docs/directories/user-directories.md)). With it off, a self-service signup creates an Active account immediately, as before. The setting is on by default and does not change anything for accounts that already exist.

In a directory that has [passwords switched off](https://dts-identity.com/docs/directories/user-directories.md#turning-passwords-off-password-optional-directories) the gate is **mandatory** and cannot be turned off, and the signup form asks for no password: the verified address is the account's only way in.

### Asking for the email address first

Instance-wide, registration can be split so the first screen asks for nothing but the email address and the profile form (name, password, custom attributes) appears behind the verification link. The account is then created when that form is submitted, still with the address already verified. Enable it with **Ask for the email address first** under [Settings → Registration](https://dts-identity.com/docs/settings/registration.md#ask-for-the-email-address-first); it does not change what the directory collects, only when.

Because the first step no longer carries a profile, it also answers identically whether or not the address is already registered: a registered address gets an email pointing to sign-in and password reset instead of a registration link, and the page shown is the same either way. A [reserved mail domain](https://dts-identity.com/docs/directories/user-directories.md#mail-domain-reservation) is the exception and is rejected on the spot, because that concerns the domain, not any individual address, and waiting for a mail that will never arrive helps nobody. Verification is inherent to this flow, so **Require email verification** has no effect on it. Invitations keep the regular one-step form.

## Account states

A user account is always in one of these **statuses**. The status names are the console labels and at the same time the values the REST API returns and accepts as `status`:

| Status | Meaning | Can sign in? |
|---|---|---|
| **Staged** | Created but not yet activated (no credentials, "activate later"). | No |
| **Provisioned** | Activation pending — an activation link was sent and is awaiting the user. | No |
| **Active** | Normal, usable account. | Yes |
| **Suspended** | Deactivated; access removed but the record retained. | No |
| **Deprovisioned** | Offboarded; assignments removed and scheduled for deletion. | No |

In addition to status, an active account can be **locked** — typically after too many failed sign-in attempts. Locking is independent of status: a locked active account cannot sign in until it is unlocked.

### Suspend vs deprovision vs delete

Before you click, know what each action keeps versus destroys:

| Action | Access | Credentials | Assignments | Record |
|---|---|---|---|---|
| **Suspend** | Blocked immediately | Kept | Kept | Kept — re-activate to restore as-is |
| **Deprovision** | Removed | Cleared | App access removed | Kept, then scheduled for deletion after a grace period (about 30 days) |
| **Delete** | Gone | Gone | Gone | **Permanent** — the record is destroyed |

Suspend is the reversible "pause" — assignments survive so you can reinstate the account. Deprovision is the offboarding step: it strips app access and credentials but keeps the record for the grace window so you can audit or reverse it before it is purged. Delete is final.

> **Note:** An account with the **owner** access level can never be deleted from the console. A **synced** user (from an enterprise directory) is deletable only once it is suspended or deprovisioned, depending on the directory's sync settings. Status changes, deletion and directory moves on an account that holds an access level also require a higher level than the target's (see [Admin Console users](https://dts-identity.com/docs/security/admin-console-users.md)).

### Changing status and acting on a user

From the user's detail page you can:

- **Change status** — for example activate a staged user, **suspend** an account (removing access while keeping the record), or **deprovision** it. Deprovisioning removes the user's assignments and schedules the record for deletion after a grace period (about 30 days).
- **Unlock** a locked account.
- **Reset MFA** — remove all of the user's enrolled second factors so they can re-enroll. This requires the user-security permission.
- **Change password** / **send activation link** — set a new password, or (re)send the email that lets the user activate and choose their own. Each send issues a new link, an earlier one stops working.
- **Verify email** or **verify mobile** manually.
- **Change directory** — move the user to another directory of the same business type (you cannot do this to your own account).
- **Reset upstream sync** — clear the pending change marker for users whose profile changes sync back to an enterprise source.

> **Activation without a password.** In a directory that has [passwords switched off](https://dts-identity.com/docs/directories/user-directories.md#turning-passwords-off-password-optional-directories), creating a user offers no **Activate now and set password** option, so the account is created as **Provisioned** and an activation link is sent. Opening that link **confirms the account** instead of asking for a password and moves it to **Active**; from then on the user signs in with an emailed code. The password rows and password actions are hidden for such users, both on the detail page and in the end-user portal.

> **Note:** For users that come from a synchronized enterprise directory, some actions are constrained — for example, a synced user may only be deletable once suspended or deprovisioned, depending on the directory's sync settings. The console enforces these guards and explains why an action is unavailable. See [User directories → Synchronization](https://dts-identity.com/docs/directories/user-directories.md).

Every lifecycle action is written to the [audit log](https://dts-identity.com/docs/monitoring/audit-logs.md) with the acting administrator recorded.

### Admin-set locks vs automatic security locks

Two different things can stop a user from signing in, and they clear in different ways:

- **Admin-set states** — **locked** (set by you), **suspended**, and **deprovisioned** persist until an administrator lifts them. Use **Unlock** or **Change status** on the detail page.
- **Automatic security locks** applied by [threat protection](https://dts-identity.com/docs/security/threat-protection.md) — a **brute-force account lockout** (too many failed attempts on one account) and a **breached-password** block (the account's password appeared in a known breach). These are not cleared by an admin status change: a brute-force lockout clears on its own once the window passes (or when the user successfully resets their password), and a breached-password block clears when the user sets a new, uncompromised password. The matching IP-level blocks are managed on the [blocked IPs](https://dts-identity.com/docs/monitoring/rate-limits-and-blocked-ips.md) page, not here.

If a user reports being locked out, check which kind it is first — an admin lock you can lift directly, an automatic one usually resolves through a password reset.

## Posture risk score

Each **active** user carries a **posture risk score** from 0 to 100 that answers *"how exposed is this account if it were compromised?"* It is not a sign-in anomaly score — it summarizes signals already on the account.

The score combines two axes, multiplied together:

- **Impact (blast radius)** — how much damage the account could do: its access level, whether it is an [organization](https://dts-identity.com/docs/directories/organizations.md) administrator, whether it holds a privileged system role, and how many applications it can reach.
- **Likelihood (exposure)** — how attackable it is: no MFA or only a weak factor, a breached or stale password, locked / failed-login state, a dormant account (no login in over 90 days), or an unverified email.

The multiplicative model means a low-privilege account with weak protection stays low, while a **privileged** account with weak protection becomes **critical**. Scores fall into four bands:

| Band | Score |
|---|---|
| Low | below 25 |
| Moderate | 25 – 49 |
| High | 50 – 74 |
| Critical | 75 and above |

The user's profile shows the band as a qualitative summary of each axis (for example, "high damage potential, well protected"), and the [People list](https://dts-identity.com/docs/directories/people.md) can sort and filter by risk so you can triage the most exposed accounts first. A [dashboard tile](https://dts-identity.com/docs/getting-started/admin-console-tour.md#activity-and-summary-tiles) shows the current number of accounts in the high or critical band.

To act on the score, add a **user risk** condition to an [authentication policy](https://dts-identity.com/docs/security/conditional-access/authentication-policies.md) rule: for example, require phishing-resistant MFA for accounts in the high band or above, or deny access to critical accounts. The policy evaluates the score at every sign-in.

> **Note:** The risk score and the risk column are only visible to roles that can read user-security information — auditors, support, operators, admins, and owners. Other roles do not see it. Non-active accounts are not scored.

## Impersonation (switch organization view)

An authorized administrator can **switch organization view** — scope the console to one [organization](https://dts-identity.com/docs/directories/organizations.md) and see that customer's data as their own admin would — to reproduce a problem or verify access from the customer's side. While the view is switched, a clearly visible indicator marks the session, and you can **end** it at any time to return to your own context. Only admins with the right access level can use it; it is gated on Customer Identity.

An application can send you into the switch-view with a return address. Both actions on that page then lead back to the application: switching the view, and leaving it again. Leaving the view from the indicator in the header instead returns you to your app list. A return address is only honored when its host matches a redirect URI of one of your registered applications.

The action is **fully audited**: starting and ending are recorded in the [audit log](https://dts-identity.com/docs/monitoring/audit-logs.md) as **Impersonation Started** and **Impersonation Ended**, with the real administrator distinguished from the organization being viewed — so the activity is always attributable. See [Organizations](https://dts-identity.com/docs/directories/organizations.md) for the full walkthrough.

## Sessions

A single sign-in spans more than one session, and it helps to keep the layers straight:

- The **platform session** at the authorization server is what drives single sign-on (SSO) across all your apps. Ending it ends SSO.
- Each **connected app** keeps its **own** session once it has received a token, governed by that app's own timeout.
- For **brokered logins** (signing in through an external IdP via a [connection](https://dts-identity.com/docs/integrations/connections.md)), the **upstream IdP** also keeps its own session — ending the platform session here does **not** end the upstream one.

So signing a user out of the platform stops new SSO, but already-issued app sessions and the upstream IdP session live on under their own rules.

### Timeouts and revocation

- **Idle and absolute (require-login-after) timeouts** for the platform session are set **instance-wide** under **Settings → Login**, not per directory. A background process also sweeps expired and idle sessions automatically and records the termination in the audit log.
- **Reviewing and revoking** — users manage their own active sessions and trusted browsers from their **account portal**, where they can review where they are signed in and sign out individual sessions or all of them at once.
- **Automatic termination** is triggered by an explicit logout, lifetime or idle expiry, a **password or email change**, and the background offline-expiry sweep.

Because session details and recent activity per session are tracked, both the user and you (through the audit trail) can see when and from where an account was active.

## See also

- [People](https://dts-identity.com/docs/directories/people.md) — the user list and detail page where these actions live.
- [User directories](https://dts-identity.com/docs/directories/user-directories.md) — sync constraints, password policy, and the account-lockout threshold.
- [Audit logs](https://dts-identity.com/docs/monitoring/audit-logs.md) — where every lifecycle action is recorded.
- [Authentication levels](https://dts-identity.com/docs/reference/authentication-levels.md) — how MFA strength maps to assurance.
