Posts on research
3 posts on research. Notes from public sources and my own test lab.
Does AI-powered vulnerability hunting actually work?
I tested AI-driven vulnerability scanning tools in my own lab environment; I'm sharing what I learned in postmortem format.
September 1, 2026·3 dk readai-securitypentestresearch
Prompt Injection Cannot Be Solved With Classical Security Models
Prompt injection is not a parsing error; it is the natural consequence of an LLM's inability to distinguish between instruction and data, and therefore a permanent patch should not be expected.
August 19, 2026·3 dk readai-securityresearchtools
Görünmez Duvarları Yıkmak: IDOR’un Sessiz Çığlığı ve Mantık Hatalarının Anatomisi
Selam dostum, bugün seninle biraz dertleşelim, biraz da ellerimizi kirletelim. Kahveni (veya o meşhur enerji içeceğini) aldıysan konuya direkt dalıyorum. Bak,...
February 8, 2026·5 dk readencryptionresearch