Such a policy should be founded on the following principles: > Transparency and clear articulation of the company’s internal use cases for generative AI; > Assessment of the ethical, social, and legal risks associated with these technologies; > Integration of principles of fairness, non-discrimination and privacy; > Establishment of internal accountability mechanisms (audit, oversight); > Annual disclosure of progress achieved, to be included in the ESG report or an equivalent publication. This policy should aim to provide a clear framework for the use of generative AI within the corporation’s ecosystem of companies or in the course of its internal processes (for instance, human resources, compliance, communications, data processing). Ever since the adoption of the European AI Act and the advancement of Bill C-27 (the Artificial Intelligence and Data Act) in Canada, companies that use or invest in AI-driven solutions must demonstrate: > that they assess risks related to discrimination, content manipulation, algorithmic biases, and confidentiality; and > that they maintain internal mechanisms for ongoing monitoring and course-correcting. Adhering to a credible voluntary code would enable Power Corporation to: > safeguard its reputation and ESG notation; > pre-empt criticism and potential litigation associated with opaque uses of AI; and > demonstrate proactive leadership, at a very modest cost. As an influential holding company, Power Corporation: > is exposed to AI systems across several of its subsidiaries (insurance, asset management, digital financial services); > could be indirectly impacted by technological misuses (algorithmic recruitment, generated content, automated fraud, etc.); > possesses both the structure and the leverage required to adopt robust and coherent internal standards. A well-defined policy on generative AI is now as essential as a policy on cybersecurity. This proposal received the support of 4.59% of the votes of holders of “Subordinate Voting Shares” at last year’s annual meeting.