Today, there are still API security threats that most WAFs and Advanced Bot Protection solutions cannot manage. In this post, we’ll explain these new types of threats and make some recommendations for features you need within solutions to protect your APIs.
When a bad actor makes a completely valid API call that evades detection by traditional security methods, this is a problem that’s not addressed by all Bot Protection and WAF solutions. In this attack instance, typical API Access management control cannot stop an attacker from leveraging an authorized session. This is because typical schema validation cannot detect API calls with completely conforming object payloads. In this case, the attacker, using a valid API call, tricks the application into returning user data and affects a breach.
In practical terms, here’s how it works. There is an instance where a misimplementation of a small application feature enables an attacker (as “user A”) to log in. Through some careful manipulation, “user A” can change the input parameters to reference “user B” and the attacker can perform actions and even extract data as if the request were coming from “user B”, though it is really coming from a valid “user A” login session. You can see where this would confound conventional security mechanisms. As the use of APIs increases, these attack situations become more common.
API security threats present themselves in two types; known attack patterns like injections (e.g., a log4j injection in an API input), and unknown attack patterns targeting application-specific business logics and data structure (e.g., Broken Object Authorization API calls). The latter calls for a different approach than that which traditional solutions can provide.
The benefits of a full stack of Web and API security solutions
A10: Insufficient Logging and Monitoring
A stack of Web and API security solutions enables organizations to sufficiently log and monitor data without slowing down performance and creating a barrier to application development and deployment.
A9: Improper Asset Management
The data within APIs have as much asset value as the APIs themselves, connecting API security with data security. A stack of Web and API security solutions allows for complete visibility into both the APIs and the data woven into them and enables organizations to apply security policy holistically.
A stack of Web and API security solutions provides an up-to-date Web Application Firewall (WAF) to manage code injections and zero-day vulnerabilities that can affect API security.
A7: Security Misconfiguration
Organizations tend to overlook the misconfiguration of the design of the scanning and testing of APIs. Oftentimes, API security testing and verification are as good as the schema. Unfortunately, it is often the case that schema is overlooked as a documentation practice. It is important for organizations to find ways to monitor existing APIs usage and double-verify API schema hand-generated from the code and not simply trust it as being effective. A stack of Web and API security solutions helps.
A6: Mass Assignment
In a business logic attack where, for example, a bad actor tricks an application into executing a whole update or whole command as an admin, how can a generic detection tool or mechanism create an app-specific profile that’s not dependent on the DevOps team’s input? A full stack of Web and API security solutions automates the activity to figure out how the app communicates.
A5: Broken Function Level Authorization
Developers design applications to protect critical functionality from unauthorized external users, and their principal concern is detecting unauthorized “north-south” API calls that facilitate communication between external entities and the applications. When developers compose services like “back end for front end” designs where the back end is being accessed by a trustworthy front end, the APIs use “east-west” calls to communicate internally. A full stack of Web and API security solutions can detect when your front end is relaying an unauthorized function call.
A4: Lack of Resource and Rate Limiting
Traditional API security solutions count API calls, but volume is not the most important consideration. API call type is also important, so organizations also use advanced bot protection that differentiates not only human to bot but also good bot to bad bot for additional security.
A3: Excessive Data Exposure
Monitoring API call volume is less critical than being able to detect if sensitive data is being exposed and in what context. A full stack of Web and API security solutions has the capacity to identify and classify sensitive data.
A2: Broken Authentication
Many API gateways provide strict token validation practices but random tokens can be generated as a result of account takeover (ATO). Oftentimes, bad actors just take over the account and use it to generate valid API tokens, then they can automate subsequent attacks to exfiltrate data. A full stack of Web and API security solutions goes up stream to ensure that you can protect the account properly.
A1: Broken Object Level Authorization
These are obvious attacks, but how do you automate detecting them? Focusing on data objects’ relationship to one another is critical. Remember our example of “User A” tricking the application into functioning as if “User A” were “User B”? An automated security solution can detect this anomaly where traditional approaches have failed.
API Security starts with discovery
As APIs feature an application-specific data layer, automated data discovery is necessary. This discovery process must feature the ability to identify sensitive data. The API inventory is the foundation for the detection and remediation of security incidents and it must be updated automatically to function the way it is supposed to.
Join us on May 3, 2002, for the webinar, API Security: Developing a Strategy That Keeps Pace With Your Business. Lebin Cheng, Head of API Security, Imperva and Amy DeMartine, VP, Research Director, Forrester will help you better understand how enterprises approach APIs and API security and provide meaningful insight and guidance. Specific learning objectives include:
- Gaining a global insight into the challenges that developers, security, and product teams face regarding APIs
- Get key recommendations on how to automatically secure APIs
- Find out what to look for from a vendor in an API Security solution, based on new Forrester research
This session will include robust Q&A participation, so spots are limited. Reserve your spot today.
Try Imperva for Free
Protect your business for 30 days on Imperva.