The Trust Gap Behind the AI Coding Boom: What 200 Security Practitioners Just Told Us

2026-04-28 · Source: ProjectDiscovery Blog · Field: Technology & Digital — Cybersecurity & Data Privacy, Artificial Intelligence & Machine Learning · Depth: Intermediate, medium

Summary

New research from ProjectDiscovery, based on a March 2026 blind survey of 200 cybersecurity practitioners in North America and Western Europe, reveals a significant "trust gap" in AI-assisted coding. The survey found 100% of respondents reported faster engineering delivery in the last twelve months, with 49% attributing this to AI. However, only 38% of security teams are comfortably keeping pace. AI coding amplifies context-heavy vulnerabilities like secrets exposure (78%), insecure dependency usage (73%), and business logic flaws (72%). A critical finding is that 66% of practitioners spend over half their week on manual validation and reproduction of findings, driven by the need to prove exploitability (59%) and address false positives (53%) from existing tools like SCA (74%) and SAST (60%). Despite these challenges, practitioners are open to AI-driven security tools for targeted testing and validation, provided they are auditable and operate within defined bounds.

Key takeaway

For Application Security Engineers struggling with the deluge of AI-generated code, your current tools are likely creating more manual validation work than value. You should prioritize adopting AI security platforms that specifically address context-heavy vulnerabilities like secrets exposure and business logic flaws. Focus on solutions that automate exploitability proof and reduce false positives. Ensure these platforms provide auditable processes and operate within defined security boundaries to rebuild trust and scale your efforts.

Key insights

AI-accelerated development outpaces security's manual validation, amplifying complex vulnerabilities.

Principles

• AI coding significantly boosts engineering delivery speed.
• Context-heavy bugs are amplified by AI-generated code.
• Manual validation consumes two-thirds of security team time.

In practice

• Focus on secrets exposure and supply chain risks.
• Evaluate tools based on reducing manual validation.
• Seek AI security platforms with clear audit trails.

Topics

• AI-assisted Coding
• Application Security
• Vulnerability Management
• Security Automation
• Supply Chain Risk
• AI Security Tools

Best for: CTO, VP of Engineering/Data, AI Architect, AI Security Engineer, Security Engineer, Director of AI/ML

Related on AIssential

• Endor Labs launches free tool AURI after study finds only 10% of AI-generated code is secure — AI-generated code often lacks security, necessitating specialized tools for vulnerability detection and remediation.
• AI leaks your company’s code secrets faster than ever — AI coding tools significantly accelerate the hardcoding and leaking of sensitive secrets, creating widespread vulnerabilities.
• GhostApproval: A Trust Boundary Gap in AI Coding Assistants — AI coding assistants are vulnerable to "GhostApproval," where symlinks and UI misrepresentation bypass security controls, enabling RCE.
• AI Finds Vulnerabilities. Security Experts Find Impact. — AI accelerates initial vulnerability pattern identification but human expertise is indispensable for impact assessment and validation.
• The AI bugpocalypse is here. Now what? - Jack Cable, Corridor — Advanced AI models are accelerating both vulnerability discovery and introduction, necessitating a fundamental "secure by design" shift in software development.
• Code Risk Intelligence: Securing AI Coding at Scale in Real Time — AI-assisted coding necessitates embedding security intelligence directly into the developer workflow to manage new risks.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by ProjectDiscovery Blog.