SHASAI is an EU-funded research and innovation project focused on the security of AI systems throughout their lifecycle, addressing both hardware and software vulnerabilities. It develops a multi-layered security framework that encompasses design, development, deployment, and operation. The project integrates security-by-design methodologies, AI supply chain protection, red and blue teams for threat testing, intelligent firewalls, and continuous security assessment. It also addresses hardware-level threats, such as side-channel attacks and fault injection. SHASAI incorporates automated security checks into CI/CD workflows and leverages virtualization platforms for threat simulation. The expected outcome is a robust, adaptable, and reliable security architecture that ensures AI systems remain resilient, traceable, and compliant with constantly evolving cybersecurity standards in high-risk environments.