Mike Vizard summarizes a Fastly survey showing rapid AI adoption for DevSecOps, detailing the integration level, trust factors, and challenges as organizations automate application security.

Survey Reveals Rapid AI Adoption to Strengthen DevSecOps Practices

Author: Mike Vizard

A global survey conducted by Fastly, involving 1,015 IT professionals with responsibilities in application security, finds that AI is playing a significant and growing role in DevSecOps practices worldwide. According to the survey:

  • 77% of respondents are already using artificial intelligence in their DevSecOps pipelines, with an additional 13% considering its adoption.
  • 25% have fully integrated AI into development workflows, and 39% have partial integration. 31% are still experimenting, while only 6% are not using AI at all.

Impact on Application Security

  • A substantial number of organizations are automating their security processes with AI. Notably, 33% said that at least half of the application security issues identified by AI tools are acted upon without human review.
  • However, full automation remains uncommon, with only 4% reporting that 76–100% of issues are resolved without human intervention. Most organizations use human oversight as an additional checkpoint.
  • Review mechanisms such as checkpoints (66%), model vetting (49%), auditing and logging (46%), and secure sandboxing (32%) are widely employed to maintain trust and security.

Trust and Challenges

  • Just 22% rated their trust in AI tools as ‘excellent,’ while 48% found them ‘good enough.’ Combined, 30% held more skeptical views.
  • False positives were a persistent issue: 49% experienced occasional or frequent false positive alerts from AI-powered tools.
  • Major challenges include integration complexity (46%), internal skills gaps (38%), distrust in AI results (36%), poor explanation of findings (23%), and compliance concerns (33%).

Benefits Reported

  • Respondents noted reductions in manual security work (55%), faster detection (50%) and remediation of vulnerabilities (36%), and enhanced triage (43%).

Strategic Focus

  • 31% reported formal efforts to explore new AI use cases within security, while 49% are informally investigating further applications.
  • The article notes that as more code is produced using AI-assisted tools, AI’s role in remediating code issues is expected to expand.

Conclusion

Organizations are accelerating their use of AI for DevSecOps and application security, but still exercise caution and include multiple review and security measures to address trust and effectiveness. Both automation benefits and ongoing skepticism highlight the need for balanced adoption strategies as AI capabilities evolve.

For the full survey report and additional context, visit the Fastly blog.

This post appeared first on “DevOps Blog”. Read the entire article here