> ## Documentation Index
> Fetch the complete documentation index at: https://docs.blend.money/llms.txt
> Use this file to discover all available pages before exploring further.

# Compliance Controls

> Sanctions screening and a per-account audit trail run on every Blend account. What they do, when they run, and what they can never do.

You cannot screen one user inside a pooled vault. You cannot export one user's history from a shared ledger. Compliance teams get nothing to hold on to.

Blend runs one account per user, so every control attaches per user. Two controls run on every account today: automated sanctions screening and a per-account audit trail.

## What per-account control changes

|                              | **Pooled products**                                     | **Blend**                                          |
| :--------------------------- | :------------------------------------------------------ | :------------------------------------------------- |
| **Screening**                | The pool has one address. Users are invisible on-chain. | **Every user's account is screened individually.** |
| **Records**                  | One ledger for everyone                                 | **A complete, exportable log per account**         |
| **Acting on a flagged user** | Touch the pool, touch everyone                          | **Block one account. No one else is affected.**    |
| **Verifying it happened**    | Trust the operator                                      | **On-chain activity is verifiable by anyone.**     |

## Sanctions screening

Every account address is screened against OFAC and global sanctions lists. Screening is automated and runs at three points.

| When                        | What gets screened                                                  |
| :-------------------------- | :------------------------------------------------------------------ |
| **Before account creation** | The user's wallet address, before a Safe is deployed                |
| **On deposit flows**        | The depositing address, on every deposit quote                      |
| **Ongoing**                 | Every account, on a recurring basis. Stale results get re-screened. |

**What a flag does.** A flagged address is blocked from Blend-routed operations. Deposits, routing, and rebalancing stop. The flag holds until a person reviews it. Flags never clear automatically. Every screening event is recorded.

**What happens if screening goes down.** The check fails closed. If the screening service is unreachable, operations are blocked, not waved through.

**What a flag can never do.** A block gates Blend's rails. It does not confiscate. Assets stay in the user's Safe, where Blend structurally cannot seize them. Non-custody is not suspended for flagged accounts. It cannot be suspended for anyone.

<Note>
  For implementation details (screening provider, caching, fail-closed behavior), see the [security deep dive](/architecture/security-deep-dive#sanctions-screening).
</Note>

## Per-account audit trail

Every action on every account is logged and exportable. Positions, flows, caps, and exception handling all operate at the account level.

* **In the portal.** Review and export any account's log from [Audit Logs](/build/portal/operations). Filter by account, action, or time range.
* **On-chain.** Every Safe is a real contract on a public network. Anyone can verify its activity on a block explorer, without asking Blend.

Your auditors get per-user records. Your regulators get an exportable trail. You get both without building either.

## Who does what

Blend's controls cover the account layer. Your compliance program covers your customers.

| You (platform)                   | Blend                                         |
| :------------------------------- | :-------------------------------------------- |
| KYC and AML for your users       | Sanctions screening on every account          |
| Licensing and customer terms     | Per-account audit trail                       |
| Your regulatory perimeter        | Structural fund segregation                   |
| Deciding who can open an account | Blocking flagged addresses from Blend's rails |

This is a strong starting point, not a compliance program. You keep yours. Blend makes the account layer visible, screened, and provable.

<CardGroup cols={2}>
  <Card title="Regulatory Alignment" icon="gavel" href="/blend/regulatory-alignment">
    How the structure maps to MiCA, US guidance, and FATF.
  </Card>

  <Card title="Accounts & Custody" icon="vault" href="/blend/account-model">
    Why per-user accounts make these controls possible.
  </Card>

  <Card title="Security" icon="lock" href="/blend/security">
    The defense layers protecting every account.
  </Card>

  <Card title="Audits" icon="file-contract" href="/resources/audits">
    11 independent audits. Every report public.
  </Card>
</CardGroup>