Survey: Privacy and Production Safety Outrank Quality Concerns in AI Pentesting

· Original article ↗

Summary

Intruder's survey of 500 US and UK security and engineering leaders found 55% use AI pentesting, while privacy (40%) and production safety (36%) were leading adoption concerns. Respondents also reported gaps between software releases and pentests.

Key points

  • Intruder surveyed 500 security and engineering leaders across the US and UK.
  • While 68% of organizations deploy code at least daily, 47% reported going more than a month between full pentests of their most frequently tested web application or API.
  • Only 4% of security leaders said their current testing frequency provides sufficient coverage; 42% said they ship code without additional security checks between pentests.
  • The survey found 55% of organizations use AI pentesting across their environment, and another 36% have tried it but not adopted it at scale.
  • Respondents most often cited data privacy (40%) and production-system safety (36%) as AI pentesting adoption barriers; 25% named testing quality as a top concern.
  • More than half of teams (56%) said they spend at least six hours per week validating security-tool findings.
  • The article recommends starting in staging, comparing AI pentest results with a manual test, reviewing evidence of what was tested, and checking how code and data are handled.

Article Details

Publisher
Intruder (survey); The Hacker News (publication)
Scope
Survey findings on AI pentesting adoption, testing practices, and security leaders’ concerns.
Sample Size
500 security and engineering leaders across the United States and United Kingdom
Key Statistics
  • 55% of surveyed organizations use AI pentesting across their environment; a further 36% have tried it but have not adopted it at scale.
  • 42% of surveyed leaders say their organizations ship code without additional security checks between pentests.
  • The leading reported AI pentesting adoption barriers were data privacy (40%), safety on production systems (36%), and auditor or regulator acceptance (28%).
  • 56% of surveyed teams spend six or more hours per week validating findings from security tooling.
  • 90% of surveyed security and engineering leaders agree that the traditional annual pentest is dead.
Recommendations
  • Begin AI pentesting in a staging environment that mirrors production.
  • Run an AI pentest alongside a manual pentest and compare the results.
  • Look for findings that include proof-of-concept code and a record of what was tested.
  • Check how code and data will be handled before granting an AI pentesting tool access.

Vendors

Countries

Related Articles