About Skyld: Pioneering AI Model Protection

"In a world where artificial intelligence is used in every sector, especially critical sectors (energy, healthcare, defense, mobility, industry,… ) We believe that it is critical that AI algorithms are secured. "

- Marie Paindavoine

Here’s what drives us…

Skyld is a deeptech startup founded in July 2023 by Marie Paindavoine, a cryptography PhD with over ten years of experience in cybersecurity. Her background led her to an early interest in AI security, notably through her expertise in homomorphic encryption an advanced technology that enables computations on encrypted data without ever revealing it, a key challenge in Privacy Preserving Machine Learning.

Today, AI is everywhere: in energy, healthcare, industry, defense, and even cybersecurity itself. But as it becomes a major driver of innovation, AI also becomes a prime target. Traditional protections are no longer sufficient against sophisticated attacks that can, among other things, extract or hijack AI models.

Our mission is clear: to develop dedicated security solutions for AI protecting against theft, reverse engineering, and adversarial examples.

While AI adoption accelerates, AI security is still left behind, creating a dangerous cyber debt. At Skyld, we believe in AI security by design.

Supported by Inria Startup Studio and accelerated by Berkeley SkyDeck, Skyld is establishing itself as a key player in securing AI models deployed in unsecured environments, ensuring integrity, confidentiality, and monetization control.