Why DevOps Fails At Application Security

As the pace of application development speeds up, how are enterprises ensuring those comprehensive security needs are being met? Despite the allocation of significant budget to the development of applications, and the continued pressure to release and update applications quickly, enterprises are still unable to secure their applications against attacks.

In a perfect world, applications would always be coded securely, pass all vulnerability scans and penetration tests, and never encounter zero-day attacks. Unfortunately, we know all too well that there’s no such thing as invulnerable code. In today’s world of rapid software release cycles, the remediation of legacy applications is usually regarded by application development teams as a tiresome and cumbersome task of questionable value, and one that slows down the pace of business and innovation.

At Prevoty, we surveyed more than 200 application and software developers to gain a better understanding of the pressures they face to release applications. We found that more than 70 percent of the respondents admit that this pressure to release updates often override security concerns, and 85 percent said that vulnerability remediation has an impact on the ability to get products and features out on schedule and on budget.

Due to the increasing pressures to release applications, update features, and fix bugs quickly, more than half of survey respondents also said they have a release cycle of one week or less, regardless of development workflow methodologies such as agile, SCRUM, Crystal, etc. Furthermore, nearly 80 percent of developers polled worry that their clients won’t trust applications if they admit there is a security flaw.

The evolving and amorphous nature of application attacks also poses a unique threat, and leaves developers without the ability to architect applications in ways that would prevent future attacks. After all, developers aren’t hackers by nature, so how do they stay ahead of never-before-seen attacks or new malware created by the hacking community? Application developers are in the business of building technology – not breaking it – so they’re not usually looking at code through the lens of a hacker.

To try and combat this while still meeting short release cycles, development teams have adapted their security practices to try to keep up with emerging threats and attacks. The practices most frequently employed are application vulnerability scans, penetration tests, and dedicated in-house security resources. Our research found that at least 82 percent of respondents say their companies perform some vulnerability scanning and/or penetration testing prior to application release.

But despite this emphasis on testing, the research also revealed something alarming: applications are still being released even before all known vulnerabilities can be fixed. More specifically, nearly half of developers admit to releasing applications with known vulnerabilities at least 80 percent of the time. That is a disturbingly high percentage of developers releasing or updating a production application, knowing that they’re vulnerable to an attack.

We’ve all seen how destructive major data breaches can be, and they seem to be increasing in frequency and scale all the time. It should be businesses’ top priority to make sure that all known vulnerabilities are being remediated. Unfortunately, at this time, the burden of the remediation process does not align developers with application security. Instead, it’s largely viewed as a threat to the overall goal of delivering business value to the organization at large, leaving too much sensitive data at risk.