October 1, 2024

API Sensitive Data Discovery: A Compliance Imperative But an Operational Nightmare

Buchi Reddy B

CEO & Founder at LEVO

API Sensitive Data Discovery: A Compliance Imperative But an Operational Nightmare

Buchi Reddy B

CEO and Founder at Levo

As a former engineering leader, I know the pressure you face to accelerate deployment without sacrificing security.

Speed is essential—faster releases push business goals forward. But as deployment rates climb, manual security checks can’t keep up, leaving your applications vulnerable.

This challenge becomes even more critical when sensitive data—like credit card numbers, addresses, PII, or PHI—flows through your APIs.

Then, a breach doesn’t just mean downtime; it means exposing customer data, where business objectives quickly derail.

Nearly half of consumers stop buying from companies they don’t trust with their privacy, and compliance violations only compound the damage.

For fintech, the stakes are even higher. PCI fines can reach $10,000 monthly until issues are resolved, and every lost PII record can cost $180.

This blog discusses achieving compliance success despite a high deployment rate.

APIs and Sensitive Data: A strong yet imperfect match

APIs are the backbone of data exchange between clients and databases, handling everything from user information to sensitive data like credit card numbers and healthcare details. Given this role, most compliance schemes—whether PCI DSS, HIPAA, or GDPR—now mandate a comprehensive inventory of APIs and a classification of the sensitive data they manage.

However, protecting and detecting these sensitive data flows is no small feat, as evidenced by the 31% of enterprises that reported sensitive data exposure in their production APIs.

For fintech companies, for example, PCI compliance requires strict protection of cardholder data. This means tracking API calls and mapping the flow of sensitive information across networks.

With the PCI DSS Phase 2 deadline approaching in March 2025, enterprises have just two quarters to meet foundational requirements like securing cardholder data, enforcing role-based access, and monitoring encryption, all of which hinge on finding APIs that handle sensitive data.

Now, imagine trying to handle this manually.

Your InfoSec engineers would need to gather a complete API inventory, reaching out to every development team for details on the APIs they’ve worked on—most of which won’t be readily available.

In fact, 75% of enterprises report not having a complete API inventory.

Even if they collect some data, it’s unlikely to cover all environments or APIs.

Engineers would then have to sift through this incomplete information to determine which APIs are handling sensitive data, what kind of data is flowing, and whether the proper security controls are in place.

All this while API sprawl continues to grow—27% of enterprises doubled their API count in 2023 compared to 2021.

A lack of API inventory is a massive problem, but it’s just the tip of the iceberg. Detecting sensitive data flows across APIs is even more complex due to these issues:

No API Documentation
Although recommended, API Documentation is typically overlooked as teams prioritize rapid feature delivery over detailed tracking. But without documentation, identifying which of your APIs handles sensitive data is impossible and error-prone at best. A lack of documentation or poor documentation forces your security teams to manually dig into codebases or interview developers just to locate critical information, none of which is scalable or accurate, resulting in an incomplete picture. The manual documentation is hardly ever updated to keep up with APIs, making the detection unreliable.
Interlinked and Complex Microservices
Microservices architecture fragments sensitive data across independent services, each processing distinct components of an API request. This distributed structure complicates tracking as sensitive data flows across services, with any updates to a microservice potentially disrupting established data mappings. Manual data tracking across multiple services requires immense effort and is not possible to replicate for 1000s of endpoints.
Dynamic Data Handling by APIs
Your APIs don’t always handle the same data every time—they might process sensitive cardholder data in one instance and non-sensitive data in another. This dynamic behavior complicates your ability to identify when sensitive data is being handled. While flexible APIs are essential for adapting to different use cases, they result in unpredictable data flows that make static tracking unreliable. Consequently, manually tracking all permutations of data flows across APIs becomes unmanageable, particularly in large fintech environments.
Diverse Data Origins within APIs
Your APIs often pull data from various sources—internal systems, third-party services, or cloud providers—each with different types of sensitive data. As the data is spread across multiple origins, it is harder to pinpoint which APIs are handling sensitive information. These distributed systems add complexity to fintech, making manual data audits extremely difficult and time-consuming.
Third-Party/ Open Source APIs
Even though they accelerate development at zero/negligible cost, reliance on third-party or open-source APIs results in losing visibility into how your sensitive data is handled since it leaves your infrastructure. Even with careful data handling practices, you can’t guarantee how the external API processes or stores that data. Manually auditing these third-party APIs could take months, especially when documentation is limited, and still leave you with incomplete insight into how your sensitive data is handled.

Given these hurdles, with PCI DSS Phase 2 less than six months away, your DevSecOps team likely doesn’t have the bandwidth or time to handle this manually. Automating the detection of sensitive data flows is the only scalable solution.

Automated API Security Platform Levo.ai

Our platform streamlines the entire sensitive data detection and API inventory process by automating the tedious tasks your team would otherwise struggle with:

Comprehensive API Inventory: We leverage your traffic and code repositories to build the complete API inventory, including internal, external, open-source, third-party, and even inactive APIs like zombie and shadow APIs—uncovering 90-250% more APIs than you were aware.
Automatic API Documentation: We generate thorough API documentation through OpenAPI/Swagger specifications, complete with over 12 critical parameters such as version details, changelog, and request-response bodies.
Sensitive Data Mapping: Our platform automatically detects and maps all sensitive data flows through your APIs, even across third-party and partner services, ensuring no blind spots.
Security Gap Identification: We identify endpoints handling sensitive data with no or weak authentication, allowing you to address vulnerabilities before they become threats to your customers and brand.

Flexible Data Tracking: All sensitive data types are mapped at both the application and environment levels, with the added flexibility to define new data types directly through the UI.

Our trace-linking capability ensures that the data flows we detect are accurate and contain actual sensitive data rather than false positives. With your permission, we collect traces and surface them alongside sensitive data, providing you with unparalleled visibility and control.