KIbSIS: AI-based Safety for Industrial Systems

To achieve greater economic efficiency, machine users and factory operators are increasingly pursuing autonomous production. Automated systems are being integrated into machines to reduce the need for human intervention, whether on-site or remotely.

However, one major challenge remains unresolved: the automatic acknowledgement of safety devices. These devices place particularly high demands on the AI systems deployed, including the protection of the AI itself, the underlying infrastructure, and continuous network monitoring.

At present, such acknowledgements are typically carried out by human actors who decide based on contextual information. The key question addressed by KIbSIS is whether and under what conditions these safety decisions can be made reliably and trustworthily by an automated, AI-driven system.

The core challenge of this sub-project lies in the long-term protection of industrial systems and AI applications, an area that, according to current research, has not yet been adequately addressed.

KIbSIS aims to design a future-proof security architecture for Industry 4.0 focusing on:

• Post-quantum-secure and crypto-agile security and safety mechanisms for industrial applications
• Efficient vulnerability management for both hardware (HW) and software (SW) components
• Coordinated security management across the entire industrial supply chain, leveraging artificial intelligence methods and ensuring their protection.

By integrating AI-based methods into security and safety frameworks, the project seeks to increase overall security levels while enhancing automation and efficiency.

The outcomes of KIbSIS will be of particular relevance to solution providers, machine tool manufacturers, component suppliers, and service providers seeking to strengthen resilience and trust in industrial environments.

Wibu-Systems contributes to KIbSIS through its dedicated sub-project ZuSIS – Future-Proof Safety for Industrial Systems, bringing decades of expertise in software protection, licensing, and cryptography to the consortium in support of Industry 4.0.

Our specific focus areas include:

• Developing post-quantum-secure and crypto-agile security and safety mechanisms for industrial environments, including interaction between PQC cryptographic inventory and SBOM, and connecting PQC algorithms to the GTA API
• Implementing efficient internal vulnerability management for hardware and software components used within the project
• Establishing coordinated security management across the industrial supply chain to ensure consistent and traceable protection
• Securing AI applications through advanced software protection technologies, such as integrity protection using steganography within AI models and runtime integrity protection of AI applications to defend against tampering