As AI becomes embedded in core business and operational functions, safeguarding these systems requires deep collaboration between AI and cybersecurity teams. At Philip Morris International, Ray, Head of AI Security, is leading efforts to ensure security is not an afterthought—but a foundational principle across AI development. This session will share how PMI is building secure-by-design AI practices in a complex, global environment, while aligning with compliance and risk expectations.

• Building integrated workflows between AI and cybersecurity to address risks early and effectively
• Operationalising security protocols tailored to AI-specific threats across global infrastructure
• Aligning security, compliance, and innovation priorities to support responsible AI at scale

Establishing guardrails for responsible AI deployment is essential for minimizing risk and ensuring ethical outcomes. This session will cover how to design and implement content filtering mechanisms and establish safeguard protocols that prevent harmful AI behaviour, especially in sensitive or regulated environments.

·       Developing technical guardrails and content filtering for ethical AI deployment.

·       Understanding the importance of regulatory compliance in AI safety design.

·       Implementing mechanisms for proactive monitoring and control of AI outputs.