Skyld with 1.5 million euros seed

Bloomhaus Ventures announces its investment in the French deep-tech cybersecurity startup Skyld. Together with BNP, Auriga CyberVentures, Sodero and several business angels, Bloomhaus participated in a €1.5 million financing round.

Skyld, founded in 2023 by Marie Paindavoine, winner of the European Cyber Woman Award, has developed hardware-agnostic software that protects machine learning models from reverse engineering, misuse and cyberattacks - without sacrificing performance or disrupting existing workflows.

Growing demand for protection of AI models

The market for AI Trust, Risk and Security Management (AI TRiSM) is forecast to grow to 7.44 billion US dollars by 2030. This is a threefold increase compared to today. At the same time, over 70% of companies have reported security incidents involving AI systems in the last two years. The risk is increasing exponentially, particularly in the area of edge devices, where AI models are often inadequately secured.

Skyld has the potential to become a category-defining market leader in cybersecurity for AI models. Their first-mover advantage, world-class team and scalable technology clearly set them apart from competitors.

Tim Schwichtenberg, Senior Investment Manager at Bloomhaus Ventures

Skyld addresses this need with a solution that combines scalability, speed and integrity, enabling companies to safely deploy AI innovations. The new funding will be used to drive product development, acquire pilot customers and initiate international expansion.

Development & Recognition

Skyld was supported by Inria Startup Studio, accelerated by the Berkeley SkyDeck program and is now part of Cyber@Station F and the Female Founder Fellowship.

• Selection among the 12 best European security startups (ECSO 2024)
• Inclusion in the top 100 AI startups worldwide (RAISE Summit 2025)
• CEO Marie Paindavoine listed by Le Point as one of the 100 Innovators of 2025

Skyld is a deep tech startup in the field of cybersecurity for artificial intelligence. The company develops a hardware-independent software solution that protects machine learning models from misuse, reverse engineering and cyberattacks without compromising the performance of existing workflows.