VULNERABILITY DISCLOSURE
Ignoring vulnerability reports is a stupid gamble.
A researcher sends a warning. The inbox stays silent. If a breach follows, “How did this happen?” should start with what happened to that report.
VEYRSEC / SECURITY BLOG
Good questions.
Stronger websites.
AI-built websites, security claims, and the checks that matter. Field notes for founders who want evidence before reassurance.
ALL ARTICLES / 07RESEARCH. CONTEXT. PRACTICAL NEXT STEPS.
↗
SECURITY NEWS
Critical PHP flaws took Dutch websites offline. Here is what matters.
Dutch organizations took websites offline as a precaution during the PHP security emergency. What is confirmed, what has been fixed, and what website owners should do.
Research / case study
From account impersonation to a documented fix
A Critical identity flaw allowed duplicate public identities. A follow-up assessment recorded that the original registration behavior was blocked.
Research / case study
ForgeGUI: a permissions gap exposed 974,868 profiles
A historical report records anonymous read access to user-profile information. The issue was database permissions, with write and administrative access blocked.
AI WEBSITE SECURITY
Your AI says your website is secure. That isn’t a pentest.
A polished website and a confident answer can hide the same old vulnerabilities. Ask for evidence, not reassurance.
PRE-LAUNCH CHECKLIST
An AI-built website security checklist before launch.
Five practical questions to take from “it works” to a security review with clear boundaries.
PENTESTING EXPLAINED
Can AI replace a human pentester? Ask what was actually tested.
AI can support a security workflow. The important question is whether the assessment produces reliable evidence about your application.
OUR PERSPECTIVE
AI can be useful. An AI-generated security claim still needs evidence. Articles distinguish founder-reported experience from published research and avoid treating older model studies as current website failure rates.