Responsible use of artificial intelligence at PMI

While AI presents substantial opportunities to assist decision-making, enhance productivity, and accelerate innovation, it also brings responsibilities. We are advancing AI in a way that is ethical, transparent, and human-centered.

AI is likely to shape jobs and career stages going forward, including early-career roles. PMI has anticipated these developments through proactive talent and capability initiatives, including its Early Careers program, which supports the development of skills, adaptability, and responsible AI awareness across our workforce.

Moving forward, we plan to reinforce our approach to accelerate career growth, building around our existing, targeted programs and aligning them to specific objectives across the talent pipeline. Our responsible AI strategy focuses on investing in building capabilities, reskilling and upskilling employees, as well as building and deploying a strong governance structure overseeing AI and workforce management. These measures are essential to enabling sustainable AI adoption while safeguarding long-term value creation.

We closely monitor the evolving regulatory environment concerning AI and employment, making transparency and privacy protections vital to mitigate reputational, legal, and regulatory risks. Proactive disclosure of our approach to AI governance and workforce readiness can demonstrate responsible governance, build trust and help avoid negative outcomes.

Our corporate AI policy

PMI has established a company-wide AI policy, applicable to all employees and any third-party workers with access to PMI systems. The policy provides a comprehensive framework for how AI is sourced, developed, deployed, used, monitored and governed across PMI.

This policy codifies PMI’s AI Guiding Principles, which form the foundation for the responsible use of AI in PMI, and includes requirements related to oversight, accountability, risk management, and compliance with existing policies and regulations such as the EU AI Act.

What our AI policy covers

A non-exhaustive list of the policy requirements are:

Data privacy: AI systems that process personal data must comply with PMI’s internal policy on personal data protection.

Cybersecurity: AI systems must meet PMI’s reliability and security standards and undergo security review prior to deployment.

Fairness and nondiscrimination: The policy mandates that AI systems must be designed, tested, and monitored to mitigate the risk of potential bias in data, models, and outputs.

Human oversight: AI systems must allow for meaningful human oversight and control throughout their life cycle to ensure the trustworthiness of their outputs.

Transparency and explainability: Sufficient transparency and explainability for AI systems must be considered by design to inform users of the AI system.

Our responsible AI initiative

As part of our responsible AI initiative, led by a cross-functional group, we are developing standards, guidelines, and procedures to operationalize our AI policy requirements (read more here). This initiative aims to integrate responsible AI tools and frameworks into all AI projects, mitigating risks such as harmful content, bias amplification, and lack of transparency. 

In addition to the AI policy and responsible AI initiative, PMI has invested in a company-wide AI literacy program that has been available since 2024. Mandatory compliance training on our AI policy was launched globally for office-based employees in early 2026.

Measuring the impact of responsible AI implementation

We evaluate and track several financial performance indicators linked to AI initiatives, including investments, revenue impacts, cost efficiencies, and return on AI investment. These are monitored internally at project or program level. The goal is to operationalize our AI policy and establish measurable dimensions across all AI policy principles.

We invest in artificial intelligence through our AI Factory—a delivery model that is providing scalable, reusable, and trustworthy AI components. These enable teams across PMI to build and deploy AI solutions faster, with robust security and responsible AI features, at enterprise scale, and in a cost effective way. Solutions include digital platforms, infrastructure, and governance programs, with all financial tracking managed internally. While certain business areas have experienced revenue growth and measurable productivity improvements as a result of AI solutions—such as faster review cycles—we do not disclose aggregated revenue or cost efficiency figures externally.

Our commitment to continuous improvement

AI continues to evolve rapidly, and we remain committed to evolving its governance and practices accordingly, through ongoing training, enhanced standards, and collaboration across functions. As we progress toward a smoke-free future and a more digital, data-driven organization, responsible AI remains a foundational element of our long-term strategy.