Burak Yalçınkaya

wasm strapi.png

Wasmtime and WasmEdge disagree on WASI file permissions: a read-only file can come back writable. For AI-agent sandboxes, that gap is a privilege escalation.

Sandboxing AI Agents with Wasm: Are the Security Guarantees of WASI Strong Enough?

Strapi banners.png

We used AI-assisted differential fuzzing to find bugs in WebAssembly runtimes, and it found a real bug in WAMR, while also confidently flagging behavior that turned out to be valid. The experiment shows where AI is already extremely useful in security testing: generating evidence and accelerating investigation, while leaving the final judgment to human expertise.

Posts by Burak Yalçınkaya

Sandboxing AI Agents with Wasm: Are the Security Guarantees of WASI Strong Enough?

AI-powered fuzzing for WAMR: What AI Found and What Expertise Confirmed

Have critical software that has to be right? Let's talk.