In general, object level authorization checks should be included for every function that accesses a data source using an input from the user, and failure to do so comes at a great risk.
Threats to cybersecurity these days are ubiquitous and relentless. It's gotten so bad that trying to keep up with them after programs are deployed has become almost impossible. However, in this age of DevSecOps, continuous delivery, and more data paydirt than ever before, shrewd organizations are helping their developers upskill into security-aware superstars that assist in eliminating common vulnerabilities before they ever make it to production. We've addressed web vulnerabilities, plus our own Top 8 Infrastructure as Code bugs, and now it's time to get acquainted with the next big software security challenge. Are you ready?
This next series of blogs will focus on some of the worst security bugs as they relate to Application Programming Interfaces (APIs). These are so bad that they made the Open Web Application Security Project (OWASP) list of top API vulnerabilities. Given how important APIs are to modern computing infrastructures, these are critical problems that you need to keep out of your applications and programs at all costs.
A perfect example of why it's essential to use code to enforce security can be found in an examination of the broken object level authorization vulnerability. This happens when programmers fail to explicitly define which users are able to view objects and data, or provide any form of verification to view, change, or make other requests to manipulate or access objects, allowing them to modify and access objects and data through API endpoints. An API endpoint is a touchpoint, often an URL, that is utilized for communication between the API itself and another system. The ability for connectivity between apps has elevated some of the world's most beloved software, but it comes at the risk of exposing multiple endpoints if they are not air-tight.
It can also occur when coders forget or inherit properties from parent classes, without realizing that doing so also leaves out a critical verification process within their code. In general, object level authorization checks should be included for every function that accesses a data source using an input from the user.
Think you're already familiar with these, and can find, fix and eliminate an access control bug right now? Play the gamified challenge:
How did you fare? If you want to work on your score, keep reading!
Object level access control vulnerabilities allow attackers to take actions that they should not be allowed to do. This might be an action that should be reserved for administrators, like accessing or viewing sensitive data, or destroying records. In a highly secure environment, it might even mean preventing anyone from viewing records at all unless they are specifically authorized to do so.
You should keep all possible actions in mind when defining object level authorization. For example, in Java Spring API, an endpoint with a potential issue might look like this:
The API endpoint deletes orders by ID, but does not verify if this order has been made by the current logged-in user. This presents an opportunity for an attacker to exploit this loophole and delete the orders of other users.
For safe access restrictions to be properly implemented, the code would look more like this:
The access control code does not need to be overly complicated. In the case of our Java Spring API environment example, it can be fixed by tightly defining who can access objects.
First, a verification process must be implemented in order to identify who is making the request:
Next, we must ensure the object id exists and belongs to the user making the request:
And finally, we proceed to delete the object:
Keep in mind that you need to ensure that the authorization method in your code aligns with your organization's user policies and data access controls. As a way to ensure that your code is fully secure, you should carry out checks to verify that users with different permission levels have access to the data they need to perform their jobs, but are prevented from viewing or changing anything that should be restricted to them. Doing so might uncover missing object control vulnerabilities that have accidentally been overlooked.
The main takeaways from these examples are to first define every action that a user could take with an object, and then add strong access controls directly to the code. And finally, never trust the inherited parent properties to do that job or to delegate that authority elsewhere. Instead, define user permissions and actions in the code explicitly for every object type that you need to protect.
Check out the Secure Code Warrior blog pages for more insight about this vulnerability and how to protect your organization and customers from the ravages of other security flaws. You can also try a demo of the Secure Code Warrior training platform to keep all your cybersecurity skills honed and up-to-date.