Secure AI Models at the Edge, on-Premise, On-Device, on Mobile …
‍
As AI spreads across industries from healthcare to defense, traditional protections are failing: more than 70% of companies experienced AI breaches last year. Skyld’s hardware-agnostic software protects machine learning models from reverse engineering and misuse (e.g., unlimited copying or unlicensed use) while maintaining performance and integrating seamlessly into existing AI workflows. Skyld's solution is deployed across sectors such as healthcare, defense, autonomous mobility, and Industry 4.0.

Why we invested

Skyld's solution addresses a critical but still largely overlooked problem: the vulnerability of artificial intelligence models to sophisticated attacks. The global market for AI trust, risk, and security management is expected to triple to $7.44 billion by 2030. Yet more than 70% of companies suffered AI breaches in the last two years, highlighting urgent demand for Skyld’s protection technology, esp. on AI powered edge devices. While We are really impressed by the drive and domain expertise of the founder and her team. Marie Paindavoine, PhD in cryptography with over ten years of experience and many recognitions in cybersecurity. Under her leadership, in less than two years, Skyld, an Inria spin-off, has established itself as a benchmark in French DeepTech. Skyld has been accelerated by UC Berkeley and selected as one of the 12 best European startups in security in 2024 by the ECSO (European Cybersecurity Organization). With its first-mover advantage, technical moat, and global recognition, we believe, Skyld is uniquely positioned to become a category-defining leader in AI model cybersecurity.

‍

Milestones
‍
Founded in 2023

Invested in since 2025

‍

Founder

Marie Paindavoine (CEO)

‍

Category
‍

Software

‍