Brand Safety Beyond Pixels - The Hidden Vulnerabilities in the Code Behind Your Models

Oct 20 1:40 PM HST :calendar:
Audience level: Keynote

About This Talk

As AI takes over more of the advertising pipeline, LLM-generated code is slipping into the backbone of these systems - and bringing new security risks with it. In this talk, we’ll look at how hidden flaws in that code can put brand safety and privacy at risk, and what it takes to keep the AI-powered infrastructure secure.

Presenters

    Photo of Klaudia Kloc

    Klaudia Kloc

    CEO, Vidoc Security

    Klaudia Kloc is the CEO and co-founder of Vidoc Security Lab, a Silicon Valley-based startup building LLM-based tool to secure complex codebases. Before founding Vidoc, she ethically hacked companies like Microsoft, Yahoo, and Uber, uncovering critical vulnerabilities as part of her offensive security research. Nowadays her work focuses on the security risks of AI-generated code and applying large language models (LLMs) to automate code reviews at scale

    Photo of Dawid Moczadlo

    Dawid Moczadlo

    Co-Founder, Vidoc Security

    Dawid Dawid Moczadło is the Co-founder & CTO of Vidoc Security Lab, building Security Code Review for the AI era. An ethical hacker recognized by Microsoft (Top-55 Security Researcher, 2021), he previously uncovered multiple 0-days. Featured by the BBC, CBS News, The Washington Post, among other global outlets.