Quick Answer
Card testing attacks happen when bots or attackers run many low-intent payment attempts to validate stolen card details. Stripe treats this as a serious fraud-control issue because it can quickly damage authorization quality, create dispute exposure, and change how the account is monitored.
What This Signal Usually Means
The core issue is not only fraudulent attempts. It is that the merchant's current controls may no longer be separating genuine customers from attack traffic effectively enough.
What Stripe Is Likely Comparing
- attempts-per-IP and attempts-per-device
- low-value authorization bursts
- AVS, CVC, and 3DS outcomes
- traffic-source changes before the spike
- fraud disputes and decline patterns after the spike
Most Common Root Causes
- weak rate limits or challenge controls
- exposed checkout endpoints hit by bots
- poor traffic-source quality
- weak segmentation between high-risk and low-risk flows
Evidence Stripe Will Weight Most
- attack timeline and concentration by IP, device, or source
- before-and-after fraud metrics after controls changed
- 3DS, AVS, CVC, and block-rate data
- segmentation showing how attack traffic was isolated
Decision Tree
- Is the activity clustered by IP, device, or velocity pattern?
- Yes: treat it as attack structure, not random fraud.
- No: continue to traffic-quality review.
- Did the spike follow a campaign or source change?
- Yes: isolate that source immediately.
- No: continue to checkout-control review.
- Are current controls broad enough to stop bots but narrow enough to protect approval rate?
- No: redesign segmented controls.
- Yes: monitor for residual attacker adaptation.
Operational Fix Sequence
- Isolate the attack source and pattern.
- Tighten rate limits, authentication, and step-up controls for risky cohorts.
- Review traffic acquisition quality.
- Measure fraud-attempt quality and approval impact after changes.
Related Problems
- Velocity Check Failures
- High Decline Velocity
- Multiple Cards Same Device
- Failed 3DS Authentication Spike
Related Guides and Hub
FAQ
Should the merchant optimize approval rate during a card-testing attack?
No. Attack containment comes first. Approval optimization should happen only after transaction quality is stable again.