Application Security Programs: A Growing Concern Amid Expanding Backlogs
The realm of application security is facing a mounting challenge that many individuals overseeing such programs are increasingly recognizing. As each quarter progresses, the backlog of security findings expands considerably. Despite the efforts and hard work put forth by security teams, the disparity between vulnerabilities that are identified and those that are remediated continues to widen.
This prevailing sentiment is now supported by concrete data. According to Black Duck’s 2027 AI Readiness Report, an extensive analysis was conducted on approximately 65.5 million application security findings derived from a sample of around 13,100 applications across 533 organizations, examined between June 2025 and May 2026. Over this span, a stark trend was observed: in June 2025, organizations uncovered 11 new vulnerabilities for every one they successfully resolved. By May 2026, this alarming ratio had escalated to 18 to 1 and even peaked at a staggering 22 to 1 in December. In April 2026 alone, an overwhelming 10.4 million unresolved vulnerabilities were added to existing backlogs. At the conclusion of the study, it was revealed that only about 10% of these findings had been addressed.
The Changing Landscape of Exploits
For the greater part of the last decade, defenders had a certain advantage in terms of time. In 2018, the average time from vulnerability disclosure to its first exploit was a substantial 2.6 years. However, the landscape has transformed dramatically; the Zero Day Clock now indicates that this period has diminished to less than one day. Emerging technologies, particularly frontier AI models, have expedited the exploitation process. What once required specialized skills and extensive developmental time can now be executed quickly and at a fraction of the cost.
In stark contrast, the pace of remediation has failed to keep up with these rapid advancements in exploit methodologies. On average, findings resulting from static analysis (SAST) are being remedied within 28 days, while open-source vulnerabilities exposed through software composition analysis (SCA) are addressed in about 39 days. Findings identified through dynamic testing take considerably longer, averaging 316 days, while manual testing findings linger for about 232 days. This mismatch creates a precarious situation wherein every open vulnerability serves as a potential entry point for attackers.
Prioritization Challenges: A Revealing Order
One of the more unsettling insights highlighted in the report revolves around the order of remediation. In an ideal scenario, one would expect that severity would dictate the priority in addressing vulnerabilities. However, the data reveals a different reality. Critical vulnerabilities in open-source systems take a median of 58 days to resolve, compared to merely 22 days for high-severity findings. In the domain of SAST, high-severity issues are often closed more slowly than those of medium or low severity.
This phenomenon stems partly from the complexities associated with remediation. Fixing a critical flaw in an open-source component may necessitate a significant version upgrade, while addressing a low-severity issue could be as simple as applying a patch. Yet, the overarching trend suggests that when faced with an avalanche of findings, security teams are more inclined to tackle the easier fixes first, rather than adhering to a strategic prioritization based on severity.
Interestingly, addressing critical vulnerabilities as a priority could enhance overall security posture, but solely relying on severity metrics is insufficient. Analysis from Black Duck’s researchers in 2025 indicated that known exploits existed for 45% of high-severity vulnerabilities and 39% of medium-severity ones. Additionally, AI models are increasingly adept at linking lower-severity weaknesses into more complex, multistep attacks, meaning that threats considered "lower priority" should not be dismissed.
Navigating a Decreased Supply of Public Intelligence
Another layer of complexity in prioritization arises from changing external factors. In April 2026, the National Institute of Standards and Technology (NIST) announced it would focus its enrichment efforts on a narrow subset of Common Vulnerabilities and Exposures (CVEs) in the National Vulnerability Database. Consequently, many new disclosures will arrive lacking the in-depth analysis that security teams often depend upon, all while the volume of AI-driven discoveries continues to rise. Organizations relying solely on NVD scores for vulnerability triage may soon find themselves in need of alternative intelligence sources.
Successful Practices of Agile Teams
Interestingly, the report also underscores that the speed of remediation correlates more closely with operational practices than with the sheer volume of vulnerabilities. Organizations that performed SCA scans 10 to 29 times each month effectively closed open-source findings in an average of 30 days, compared to 66 days for teams scanning less frequently. This correlation highlights how seamless integration of testing into operational pipelines can facilitate quicker resolutions, with findings directed straight to developers for remediation.
Moreover, larger organizations, particularly those with more than 10,000 employees, exhibited a more disciplined approach to prioritization. They managed to close critical and high-severity vulnerabilities in a median of just 12 days, compared to 19 days for all findings combined. In contrast, several smaller groups displayed the opposite trend, taking longer to resolve their most serious vulnerabilities than those of lesser importance.
Strategic Recommendations for Practitioners
The findings from the report present a series of actionable recommendations for security practitioners:
-
Data Review: Organizations should begin by assessing their own data—are critical vulnerabilities closing faster than low-severity ones? If not, it’s essential to investigate the reasons behind this discrepancy.
-
Contextual Prioritization: It’s crucial to prioritize vulnerabilities based on exploitability, reachability, and overall business context, rather than relying solely on severity scores. Setting timelines for addressing even lower-severity findings is essential.
-
Holistic Tracking: Security teams should concurrently monitor three critical metrics: median time to closure, the percentage of findings addressed, and the overall discovery-to-closure ratio. A rapid median time is insignificant if only a small fraction of vulnerabilities ultimately gets fixed.
-
Address Coverage Gaps: The data reveals that a significant two-thirds of applications were tested using merely one methodology. Notably, manual testing—which accounted for the fewest findings—often identified critical or high-severity issues more than half the time.
- Utilize Automation: Automation should be leveraged for tasks that can be automated, including dependency upgrades and validated code fixes, allowing skilled personnel to focus on more complex judgment calls.
Ultimately, simply increasing team headcount will not resolve the staggering growth of vulnerabilities being reported monthly. This challenge is fundamentally rooted in operational models, rather than staffing levels. To effectively navigate these tumultuous waters, security teams must adopt a deliberate approach to remediation, attentive to both the urgency and complexity of findings, all while acting swiftly to mitigate exploitable vulnerabilities before attackers can exploit them.

