# Autopilot Autopilot is a [Fraud Detection Pro](https://www.checkout.com/docs/business-operations/prevent-fraud/upgrade-to-pro.md) feature that uses machine learning to screen payments before authorization. It tunes itself automatically as fraud patterns change, so you do not need to configure or monitor any rules for it. > **Information** > > To enable Autopilot, contact your account manager. --- ## How it works When you enable Autopilot, you agree on a target fraud rate with your account manager. Your fraud rate compares your fraud reports to your total sales. In the Dashboard, you can view your fraud rate by count or by amount. When you choose your target, you see the available options and how each one affects your fraud rate and acceptance rate. Autopilot then adjusts its thresholds automatically to keep you close to your target as your traffic and fraud patterns change. If Autopilot declines a payment, you receive the `43301 - Fraud score exceeds threshold` response code. For more information, see [Checkout.com-level response codes](https://www.checkout.com/docs/developer-resources/codes/api-response-codes.md#Checkout.com-level_response_codes). --- ## Autopilot and your risk strategies Autopilot works alongside your existing risk strategies. Your risk rules and lists continue to run as normal, although Autopilot generally replaces behavior-based rules. Your pre-auth risk strategy takes effect before Autopilot. If one of its rules declines a payment, the payment does not reach Autopilot for a decision. For example, this happens if the card is on one of your decline lists, or if the payment matches one of its decline rules. For more information, see [Create risk rules and lists](https://www.checkout.com/docs/business-operations/prevent-fraud/create-risk-rules-and-lists.md) and [Build and test risk strategies](https://www.checkout.com/docs/business-operations/prevent-fraud/build-and-test-risk-strategies.md). --- ## Improve Autopilot's performance Autopilot makes better decisions when it has more information about each payment. To get the best results, provide as much of the following data as possible in your payment requests: - Customer name - Customer email address - Billing address - IP address - Device data, collected using [Risk SDK](https://www.checkout.com/docs/business-operations/prevent-fraud/integrate-with-risk-sdk.md) To learn how the data you provide affects fraud prevention, see [Optimize payment request data](https://www.checkout.com/docs/business-operations/prevent-fraud/optimize-payment-request-data.md). --- ## Monitor performance You can view Autopilot's performance on the _Performance_ page in the Dashboard, alongside your other risk analytics. You need one of the following [user roles](https://www.checkout.com/docs/business-operations/use-the-dashboard/manage-users/user-permissions.md): - Admin - Risk manager - A custom role with the `Create, edit, or delete rules` permission 1. Sign in to the [Dashboard](https://dashboard.checkout.com). 2. In the business entity selector, select an entity. 3. Go to _Payments_ > _Fraud_ > _Performance_. 4. Select the period you want to analyze. In the _Blocked transactions_ section, the _Autopilot_ category shows the transactions that Autopilot blocked. For more information, see [Monitor risk performance](https://www.checkout.com/docs/business-operations/prevent-fraud/monitor-risk-performance.md#View_risk_analytics).