| [ Web Proxy ] |
| Viewing: https://docs.stripe.com/radar/risk-settings | [Back] [Original] |
If your Radar plan supports this feature, you can choose a risk setting that automatically adjusts your protection thresholds based on your business needs and tolerance. Based on the risk setting you choose, we recommend the risk controls that help protect against fraud.
Risk settings let you balance authorization and fraud on your account by using risk controls. These risk controls set different thresholds to block payments that might result in fraudulent disputes or early fraud warnings. Certain risk settings also recommend that you enable or disable additional risk controls, such as adaptive 3DS and dynamic risk thresholds, to help protect your account from fraud while allowing non-fraudulent payments.
On the risk controls page, you can choose or modify your risk setting from the options below. You might need to upgrade your plan to adjust your risk setting.
When you change your risk setting, the modal displays the impact to payments with the following statistics. The numbers shown are estimated projections based on your last four months of payment data.
| Metric | Description |
|---|---|
| Fraud | The estimated payments this setting blocks before they become fraudulent disputes or early fraud warnings. |
| Payments | The estimated payments and percentage change your business accepts, based on what this setting allows. |
Use risk settings instead of default risk thresholds to change your blocking preferences. If you previously set a risk score block threshold, the Block if :risk_level: = 'highest' default rule is likely still enabled. If so, you wont see the option to change your risk score threshold on the risk controls page.
If you want to change your default block threshold, you can do either of the following:
Risk controls use AI to protect your account. You enable them on your account to block payments that might result in fraudulent disputes or early fraud warnings. Risk settings determine which risk controls are enabled and recommend additional risk controls, such as adaptive 3DS or dynamic risk thresholds.
Risk settings automatically manage the blocking thresholds for risk controls. To block more or less traffic from a risk control, you can change your risk setting, which automatically adjusts the blocking threshold. You can also adjust the threshold for the Fraudulent dispute risk control directly.
Risk controls and risk settings wont override the custom rules you created. You can continue to update your controls, settings, and rules at any time.
You can find the following controls on the risk controls page in the Dashboard.
This risk control blocks payments that Stripe Radar determines are likely fraudulent, even when those payments are unlikely to result in a dispute or early fraud warning. Rather than waiting for issuer reports or chargebacks, it acts at the time of payment, stopping fraud before it completes.
Radar determines the blocking threshold based on your risk setting. You can monitor performance and review affected payments on the risk controls page in the Dashboard. You can also backtest this control to simulate how it would have affected past payments and understand the impact on conversion and fraud prevention.
Fraudulent card payments is included with Radar Standard and above.
This risk control uses the fraudulent dispute score and blocks payments that are likely to result in fraudulent disputes. Radar determines the number of payments to block versus accept based on your risk setting, and maximizes conversion while preventing the risk of entering monitoring programs.
To change how much this control blocks, change your risk setting.
You can also click Edit configuration on the controls overview page to set a custom blocking threshold. You can backtest the impact of the threshold on your payments before applying it.
This risk control uses the early fraud warning score and determines which payments are likely to result in an early fraud warning. Radar determines the threshold based on your risk preferences, and maximizes conversion while preventing the risk of entering monitoring programs.
This control is available on the Balanced and Maximize protection risk settings. We recommend turning on this setting if youre in a card brand monitoring program, which counts early fraud warnings toward the VAMP monitoring program.
Radars adaptive 3D Secure (3DS) authentication control adds an extra layer of authentication during checkout to help protect against fraud. Stripe runs machine learning in the background to authenticate medium-risk payments.
In most cases, 3DS authentication happens with no action required by your customer. If the customers bank detects anything unusual, they might ask the customer to confirm their identity through a security challenge. If 3DS authenticates a payment, the liability for any fraud-related disputes for that payment typically shifts from your business to the issuer. This means that in most cases, the seller isnt responsible for fraud costs on 3DS-authenticated payments.
Stripe activates 3DS automatically, if required by a regulatory mandate such as Strong Customer Authentication (SCA). Disabling adaptive 3DS doesnt prevent 3DS from activating when required for compliance. Authentication fraud prevention control isnt available in testing environments.
The dynamic risk thresholds control automatically blocks additional elevated and high-risk payments when your account is under fraud pressure. Stripe continuously monitors for fraud trends using machine learning and temporarily blocks these payments to help protect against fraud during a detected threat. When we no longer detect a threat, the block threshold automatically readjusts to your default setting.
Dynamic risk thresholds apply to card payments only.
This risk control operates automatically in the background. You dont need to create custom rules or adjust settings during a detected threat. Dynamic risk thresholds use your manual threshold settings if theyre configured to block more elevated and high-risk payments. You can monitor performance and review affected payments through detailed analytics in your Dashboard. This automated approach reduces your need to:
Stripe automatically enables the fraudulent non-card payments control by default for all non-card, local payment methods. This risk control evaluates each transaction and automatically blocks payments that have a high risk level. Radar determines the risk level based on attributes that Stripe has identified as high fraud risk on other payments.
You can monitor performance and review information about blocked payments on the risk controls page in the Dashboard. You can also simulate how this risk control would have affected past payments to better understand the impact on conversion and fraud prevention.
Radar scores are numerical signals that represent the likelihood of fraud or abuse for a payment. You can use scores to write custom Radar rules or to understand why a payment was blocked. Some scores also contribute to an overall risk level, which summarizes risk as a single signal.
The overall risk level combines the fraudulent dispute, early fraud warning, and risk scores that factored into the payment decision. It takes the maximum of these scores and appears as a single signal (Normal, Elevated, or Highest) on the payment details page. The overall risk level is separate from the :risk_level: Radar rules attribute, which only corresponds to the risk score. The highest applicable score also appears on the Radar Reviews page and in the Related payments section of the payment details page in the Dashboard.
Stripe Radar assigns each card payment a fraudulent payment score between 0 and 99. The score represents the likelihood that a payment is fraudulent, where 0 is the lowest likelihood and 99 is the highest. Radar determines the score based on attributes Stripe has identified as indicators of high fraud risk across the Stripe network.
Fraudulent payment scores apply to card payments only.
The Fraudulent card payments risk control uses this score to determine which payments to block. If your Radar plan supports custom rules and you want to block payments based on the fraudulent payment score, you can write custom rules on :fraudulent_payment_score:.
Stripe Radar assigns each payment a numerical fraudulent dispute score between 0 and 99. This score represents the likelihood that a payment results in a fraudulent dispute, where 0 is the lowest likelihood and 99 is the highest.
Fraudulent dispute scores apply to card, ACH Direct Debit, and SEPA Direct Debit payments.
If you want to block payments based on the fraudulent dispute score, you can write custom rules on :fraudulent_dispute_score:.
Stripe Radar assigns each payment a numerical early fraud warning score between 0 and 99. This score represents the likelihood that a payment results in an early fraud warning, where 0 is the lowest likelihood and 99 is the highest.
Early fraud warning scores apply to card payments only.
If you want to block payments based on the early fraud warning score, you can write custom rules on :early_fraud_warning_score:.
Stripe Radar assigns payments made on Stripe Checkout a numerical bot score between 0 and 99. This score represents the likelihood that a bot made the payment, where 0 is the lowest likelihood and 99 is the highest.
You can write custom rules on :bot_score: to block payments that exceed a threshold you define. For early access, see Bot abuse prevention.
Stripe Radar gives each payment a numerical risk score between 0 and 99, where 0 is the lowest risk and 99 is the highest. This risk score is the probability that a payment results in a fraudulent dispute or early fraud warning.
Adjusting the risk score applies to card, ACH, and SEPA Direct Debit payments. You can select a specific payment method to test different thresholds and view affected payments.
Radar is deprecating the high risk block rule that uses risk score to block payments. To change your block preference, you must use risk settings, which use separate models to block payments. When you select a risk setting, your default high risk block rule is disabled and you cant re-enable it.
The risk score appears on a payment when:
:risk_score: predicate or a Stripe block.The risk score factors into the overall risk level when visible.
If you want to block payments based on the risk score (even after selecting a risk setting) you can continue writing custom rules on :risk_score:.
Radar runs your Radar rules on payments made to your platform, including transactions where your platform is the merchant of record.
To apply your Radar rules on transactions where your connected account is the merchant of record, such as direct charges, do the following:
These settings apply only to connected accounts controlled solely by your platform.
If you update your connected accounts plan to Radar Lite, or choose Only my platform for managing fraud prevention rules and settings, you can also choose whether connected accounts can see Radar in their Dashboard.
Under Dashboard access, enable Make Radar visible for connected accounts to grant connected accounts view-only access to outcomes. Connected accounts cant change Radar rules or settings.
| Web Proxy Viewer | New URL | Original Page |