Skip to main content
Multi-factor proves somebody holds a credential. It cannot tell you they did not hand that credential to somebody else. Here’s how we check the same human is still there with HumanCheck:
  1. The first check sets the baseline. Run one as early as you can, at signup or when somebody completes a profile, and every later check compares against that face.
  2. Every later check asks whether this is the same person. If it is, they are recognized and pass. If somebody else is holding the account, the run does not tie back to the person who opened it, and the history shows two people where you expected one.
  3. Recognition carries across your checks. Somebody verified at signup is recognized at payout, so the second check is a few seconds.
  4. You choose the moments. A payout, a withdrawal, a high value action, or a fixed interval. Each is its own verification with its own settings.
  5. The history sits on the person. Every confirmation they appear in, newest first, with the devices we know them on and what they have proven.
This tells you whether it is the same person, not who they are. To sweep the accounts you already have once, see Userbase healthcheck.

​
How you start

  1. Create the baseline verification and put it as early in the account’s life as friction allows.
  2. Set Identity to Linked by guest, so a phone or email carries recognition when somebody changes device.
  3. Create a second verification for the recurring moment, and set its Verification limit to that cadence, for example once per period.
  4. Turn on Skip check on known devices if you would rather people on a device you already know pass without scanning.
  5. Read History on the person, which lists every confirmation they appear in.

​
Next steps

Guests and members

How recognition carries across devices and checks.

Configuration

Every setting on a check.

Comprehensive trust score

The person view, and the history behind it.

Cross-network recognition

Recognition beyond your own workspace.