Autorisk runs every charge through the rules you write. Combine fifty-plus signals from across the Worthy stack into rules that block the charge, send it to review, let it pass through, or step it up to 3D Secure. Configure each rule yourself, or let AI Companion draft them from your history. Worthy flags high-fraud incidents the moment they show up.
Real-Time Decisioning
Worthy gives you the building blocks. You decide which charges go through, which get reviewed, which get blocked, and which need a step-up. Every rule runs in real time against every charge, with the full reason returned on every response.
SIGNALS TO BUILD WITH
<
0
+
Across transaction, customer, behavioral, network, merchant, and adaptive categories.
<
0
+
RULE ACTIONS
0
Pass, trigger 3DS, review, or block the charge.
0
EVALUATION TIME
<
0
ms
Rules run before the transaction reaches authorization.
<
0
ms
Pick the action that fits the risk. Pass charges through when they look clean. Send borderline ones to the review queue. Block the obvious bad actors before authorization. Step up the suspicious ones to 3D Secure to shift liability and add an extra verification layer.
Every rule combines one or more signals from the Worthy library. Transaction signals catch obvious mismatches. Customer signals carry account history. Behavioral signals expose unusual patterns. Network signals carry intelligence from across the Worthy platform. Merchant signals tune for your category. Browse the library, mix and match, write the rules that fit how you sell.
Amount, currency, time of day, AVS match, CVV match, BIN intelligence, payment method, and currency-country alignment. The basics that catch obvious mismatches before authorization.
Account age, charge history, dispute rate, refund history, success rate, and repeat customer status. Long-term context that says whether this customer has earned your trust.
Velocity counters, device fingerprint history, session timing, geographic patterns, bot classification, and account sharing match. The patterns that show up when a real customer is not the one typing.
Cross-Merchant Fraud Network match, BIN reputation, issuer signals, identity rotation matches, and global anomaly indicators. Intelligence the rest of the Worthy platform learned for you.
Vertical-specific risk patterns, MCC baseline anomalies, your own merchant history, and category-tuned thresholds. Risk that only makes sense for the shape of your business.
Continuously refreshed signals that pick up novel patterns the static ones miss. Surface emerging fraud techniques weeks before they would otherwise be flagged in your own data.
Three example rules from a typical merchant rule set. Chain signals with AND, OR, and NOT, then route every match to one of the four actions. Build rules in the dashboard, copy them from a template library, or let AI Companion draft them from your own fraud history.
High-value,
first-time customer
Turn signals into decisions at scale
Every action Autorisk takes is recorded with the rule that produced it, the signals that triggered, and the values each signal carried at decision time. The same trace is available through the API, in the dashboard, and in every export, so your fraud team and your auditors are always looking at the same thing.
Track which rules are firing, which signals are pulling the most weight, and how your block and review rates trend day over day. Worthy auto-surfaces fraud incidents the moment one shows up, so you can adjust rules before the spike turns into a chargeback wave.
Autorisk runs automatically on every charge. The matched rule, the action taken, and every signal that triggered are attached to the charge object. Subscribe to the autorisk.matched webhook for real-time delivery to your own systems.
Read the autorisk object on every charge response.
action returns pass_through, review, require_3ds, or block.
matched_rule names the rule that produced the action.
fired_signals lists every signal that contributed, with category and value.
incident_alert appears when this charge is part of a detected fraud spike.
const response = {
"id": "ch_4PQ8K",
"autorisk": {
"action": "require_3ds",
"matched_rule": "new_customer_large_charge",
"rules_evaluated": 12,
"fired_signals": [
{
"name": "amount_gt_500",
"category": "transaction",
"value": 684
},
{
"name": "new_customer",
"category": "customer",
"value": "true"
},
{
"name": "issuer_country_mismatch",
"category": "transaction",
"value": "true"
}
],
"incident_alert": null,
"evaluation_ms": 42
}
};
Bot Detection feeds a behavioral signal you can write rules against. Cross-Merchant Fraud Network feeds a network signal. Account Sharing feeds a customer signal. Velocity Rules tightens the velocity counters Autorisk reads. Pick the building blocks, write the rules.
Confirmed bad actors blocked across every merchant. Feeds the network signals you can use in any Autorisk rule.
Cap repeat attempts by card, IP, device, or customer. Velocity counters feed straight into Autorisk rules.
Bot classification adds a behavioral signal you can write rules against to block automated traffic at checkout.
The fraud-rules engine running across the Worthy platform. Fifty-plus signals, four actions, sub-50ms evaluation.
Apply for a Worthy account and start building rules from the Autorisk library. Configure them yourself, copy from the template set, or let AI Companion draft them from your fraud history. Talk to sales for a custom package or volume pricing.
