User profiles let platforms scope Claude safety actions to one end customer

what changed

On 20 October 2026, Anthropic launched user profiles. They let you tell Anthropic which of your end customers each API request is for. If you run a platform or gateway, or resell API access, a safety enforcement action can then apply to one end customer instead of your whole organization.

The SDK and the ant CLI support the new /v1/user_profiles endpoints.

The same day, Anthropic added organization webhooks to the Admin API, in beta with the webhooks-2026-10-07 beta header. They cover two events: when Anthropic restricts or bans one of your organization’s user profiles, and when that restriction or ban ends. Access needs a request to your Anthropic account team.

Inference hooks also moved out of beta for Claude Enterprise organizations.

why it matters

If your product serves many customers through one API account, this can narrow who an enforcement action affects.

It also gives you something to build on. Webhooks tell your systems when a profile is restricted, so you can react in code instead of finding out from failing requests.

what we would do

• Map each of your end customers to a stable, non-personal identifier, and decide what you will send before you wire it in.
• Read the user profiles docs and the endpoint reference, then test on a non-production account.
• If you can get webhook access, plan the handler: stop traffic for that customer, tell your support team, and log the event.
• Add the profile identifier to your own request logs, so you can match a restriction to what that customer did.

Our build service covers this kind of production hardening.