Valliance logo in black
Valliance logo in black

"Aardvark targets the code-review bottleneck and rising risks of agentic coding. However, CTOs should see it as a mandate to modernise the SDLC as without human oversight it will amplify, not reduce, risk."

Introducing Aardvark: OpenAI’s agentic security researcher

Source Summary

Aardvark is an agentic security researcher powered by GPT-5 that analyzes code repositories to identify and help patch vulnerabilities using advanced techniques like threat modeling and commit scanning. It has been tested internally and with partners, achieving high recall on benchmark repositories and responsibly disclosing issues in open-source projects. A private beta is available for select partners.

_Related thinking
_Related thinking
_Related thinking
_Related thinking
_Explore our themes
_Explore our themes
_Explore our themes
_Explore our themes