How a Leading Online Giving Platform Shut Down Card Testing Attempts at the Edge

How one online giving platform used Darwinium to stop card testing attacks, eliminate direct API abuse, and protect the donation experience without changing its existing payment infrastructure.

Every donation should make a difference. But for online giving platforms, not every payment attempt comes from someone looking to support a good cause.

Increasingly, fraudsters are exploiting digital donation platforms to test stolen credit card details, taking advantage of low-value transactions, frictionless payment experiences, and the openness that makes online fundraising so effective.

The consequences extend far beyond unsuccessful transactions. Card testing attacks can generate significant payment processing costs, increase chargeback exposure, consume fraud investigation resources, and threaten the seamless donor experiences that charities depend on.

So how can online giving platforms protect themselves from sophisticated payment fraud without introducing unnecessary friction for legitimate donors?

In our latest case study, we reveal how a leading US-based online giving platform partnered with Darwinium to interdict nearly 70% of attempted payments, reject more than 186,000 card-testing attempts in just 90 days, and stop fraudulent transactions before they reached the payment processor.

And it did so by deploying fraud prevention directly onto its existing AWS CloudFront infrastructure, without making changes to its donation forms.

Download the Case Study: How an Online Giving Platform Shut Down Card Testing at the Edge →

Why Online Donation Platforms Are Attractive Targets for Card Testing Fraud

Online giving platforms play an essential role in connecting donors with the charities and causes they care about.

By making digital fundraising simpler, faster, and more accessible, these platforms help nonprofit organizations reach wider audiences and generate vital donations.

Unfortunately, the same frictionless payment experiences that benefit legitimate donors also create opportunities for fraudsters.

Card testing fraud occurs when criminals use online payment forms to validate stolen credit card credentials, often by attempting multiple low-value transactions.

Once a stolen card is confirmed as active, its details may be used for larger fraudulent purchases or other criminal activity.

Donation platforms are particularly attractive targets because they often allow small payments, prioritize ease of use, and aim to minimize barriers between donors and charitable contributions.

For platforms processing significant transaction volumes, even relatively small individual attacks can quickly escalate into a substantial operational and financial burden.

Every fraudulent payment attempt can potentially result in:

• Payment processing fees, even when transactions are subsequently declined.
• Increased chargeback exposure and associated costs.
• Additional operational resources spent investigating suspicious activity.
• Reputational damage and disruption to legitimate fundraising.
• Unnecessary friction for genuine donors when fraud controls become overly restrictive.

The challenge is not simply detecting fraudulent payments. It is identifying malicious intent early enough to prevent fraudsters from exploiting the payment infrastructure in the first place.

The Hidden Challenge: Card Testing Attacks That Bypass Donation Forms

For the online giving platform featured in our latest case study, card testing was already a persistent problem.

The platform had an incumbent fraud prevention solution in place, but needed to replace it quickly with an approach capable of identifying attacks earlier in the payment journey.

Initial analysis uncovered an important visibility gap.

Around 60% of attempted payments were arriving through a direct API path, bypassing the donation form entirely.

These payment attempts were not being profiled by the existing browser-based controls.

Meanwhile, attackers using the donation forms were employing increasingly sophisticated techniques to disguise their activity.

These included clearing cookies, rotating identifiers, using automated browser extensions, and deploying CAPTCHA-solving tools to evade detection.

Traditional fraud controls that rely heavily on individual transactions, browser cookies, or static device identifiers can struggle to recognize these patterns, particularly when attackers deliberately change their digital footprint.

The platform needed an approach that could:

1. Detect and stop card testing before payment attempts reached the processor.
2. Protect both browser-based donation journeys and direct payment API traffic.
3. Deploy quickly without disrupting existing donation experiences.
4. Recognize repeat attackers even when they attempted to disguise their devices or automate interactions.

The solution was to move fraud detection and decisioning closer to where the attack originated.

Stopping Card Testing Fraud at the Edge with Darwinium

Rather than relying exclusively on fraud checks within the payment application, the platform deployed Darwinium directly onto its existing AWS CloudFront content delivery network (CDN).

This enabled fraud profiling and real-time decisioning at the edge, covering both the donation forms and payment API endpoints.

By assessing activity before it reached the payment processor, the platform could identify suspicious payment attempts and take immediate action.

1. Recognizing Repeat Card Testers with Device Intelligence

Fraudsters frequently attempt to disguise their identities by clearing cookies, rotating browser storage, or modifying device attributes.

Darwinium's device intelligence uses digital signatures and similarity matching to recognize devices even when some of these identifiers change.

This allowed the platform to identify suspicious patterns of repeated payment attempts originating from the same underlying device.

In one particularly striking example, a single device signature was linked to more than 11,000 card attempts over 12 days, representing more than 20% of the platform's payment volume during that period.

Despite attempts to disguise the device, Darwinium identified the repeated activity and blocked it.

2. Detecting API Abuse That Traditional Browser Controls Missed

Browser-based fraud prevention is only effective when suspicious activity actually passes through the browser.

For this platform, a substantial proportion of payment attempts bypassed the donation form altogether.

Darwinium's edge-based deployment provided visibility into these direct API requests, enabling the platform to assess and control activity that had previously gone unprofiled.

The result?

100% of unprofiled direct payment traffic was blocked unless explicitly allow-listed.

This closed a significant visibility gap and prevented attackers from using the payment API as an alternative route around fraud controls.

3. Identifying Automation and AI-Assisted Attacks

Not every automated interaction is necessarily fraudulent.

As AI agents and automated browsing tools become more prevalent, digital businesses need to distinguish malicious automation from legitimate activity.

Darwinium uses behavioral and browser integrity signals to identify suspicious automation, including CAPTCHA solvers, browser inconsistencies, and unnatural interaction patterns.

Crucially, this intelligence enables more selective intervention rather than indiscriminately blocking all automated traffic.

In this case, only 21% of payments flagged as agent-related were blocked, compared with a 70% block rate across all payment attempts.

This demonstrates the importance of making contextual risk decisions rather than treating every automated interaction as inherently malicious.

4. Turning Fraud Detection into a Deterrent

Blocking a fraudulent payment is important. Preventing the attacker from immediately trying again is even better.

Darwinium enabled the platform to return realistic payment decline responses when card testing was detected.

Rather than revealing that sophisticated fraud controls had identified the attack, the platform could return a normal-looking decline response.

To the fraudster, the card simply appeared unsuccessful.

This approach helped reduce the value of repeated testing attempts and encouraged attackers to move on.

The Results: 186,000+ Card-Testing Attempts Rejected in 90 Days

The impact of moving fraud prevention to the edge was substantial.

Across a 90-day period, the online giving platform achieved:

• 186,063 card-testing attempts rejected
• Fraudulent attempts stopped before reaching the payment processor
• Nearly 70% of attempted payments interdicted
• Significant reduction in malicious payment activity
• 11,000+ attempts linked to one device signature
• Persistent attackers identified despite attempts to evade detection
• 100% of unprofiled direct API traffic blocked unless allow-listed
• A major fraud visibility gap closed
• No donation form changes required
• Rapid deployment using existing AWS CloudFront infrastructure

Beyond stopping fraudulent transactions, the platform avoided processing fees on rejected attempts, generating hundreds of thousands of dollars in operational savings according to the case study.

The platform also gained a stronger foundation for identifying suspicious activity across its payment infrastructure without introducing blanket restrictions on legitimate donors.

As the customer explained:

"The speed and simplicity of deployment was a real benefit to us. The path to blocking card testing was extremely slick and we feel confident we are protecting the platform from even the most persistent abuse."

— Customer, Online Giving Platform

Download the Full Case Study to Explore the Results →

Why Real-Time Fraud Prevention at the Edge Matters

This case study highlights an important shift in how digital businesses can approach payment fraud prevention.

Traditional fraud solutions often focus on individual transactions or specific moments within a customer journey.

But sophisticated fraud attacks rarely respect those boundaries.

Attackers move between browsers and APIs, rotate device identifiers, manipulate application behavior, and exploit gaps between security and fraud prevention systems.

Stopping these attacks requires more than another point-in-time fraud score.

It requires visibility into the underlying behavior and intent of every interaction, combined with the ability to act immediately.

By deploying fraud intelligence and decisioning at the edge, Darwinium enables businesses to:

• Identify malicious activity earlier: Detect suspicious behavior before it reaches payment infrastructure or backend applications.
• Uncover previously invisible attacks: Extend protection across browser journeys and APIs rather than relying solely on form-based profiling.
• Recognize persistent fraudsters: Use device signatures and behavioral intelligence to connect suspicious activity across changing identifiers.
• Respond in real time: Apply tailored pass, challenge, or reject decisions according to risk.
• Reduce operational complexity: Deploy and adapt fraud strategies through existing CDN infrastructure without extensive application engineering changes.

For businesses balancing fraud prevention with customer experience, this creates an opportunity to stop more malicious activity while preserving seamless journeys for trusted users.

Protecting Donations Without Putting Barriers in Front of Donors

For charities and nonprofit fundraising platforms, trust is everything.

Donors expect their contributions to be processed securely and efficiently. Charities rely on digital platforms to make fundraising accessible. And platform operators need to protect their payment infrastructure without undermining either of those priorities.

The challenge is finding the right balance between fraud prevention, operational efficiency, and donor experience.

Our latest customer case study demonstrates what becomes possible when payment fraud is detected and stopped before it reaches the processor.

By combining edge-based deployment, persistent device recognition, API visibility, and real-time decisioning, Darwinium helped a leading online giving platform shut down large-scale card testing while keeping its donation infrastructure running.

Want to see exactly how they did it?

Download the full case study to discover:

• How the platform deployed Darwinium on AWS CloudFront without changing its donation forms.
• How device intelligence uncovered more than 11,000 card attempts from a single device signature.
• How direct API abuse was identified and controlled.
• How real-time payment decisions stopped more than 186,000 fraudulent attempts in 90 days.
• How the platform reduced unnecessary processing costs while protecting legitimate donor journeys.

Download the Case Study: How a Leading Online Giving Platform Shut Down Card Testing at the Edge

Discover how Darwinium's Intent Intelligence platform helps digital businesses detect fraud, prevent abuse, and protect revenue by understanding user behavior and intent across every digital interaction.