BCGEU has submitted a shareholder proposal to be considered at Thomson Reuters’ (TSX/Nasdaq: TRI) annual general meeting on June 4, 2025. Our proposal asks the Company to strengthen its AI governance framework by aligning it with the UN Guiding Principles on Business and Human Rights (UNGPs) and to assess whether its longstanding Trust Principles adopted in 1941 to safeguard journalistic integrity are sufficient to govern 21st-century AI-related risks. You can read our Investor Brief here.

Artificial intelligence (AI) is a Janus-faced technology. On one side, it holds enormous potential to drive innovation, improve lives, and solve complex problems. On the other side, it raises profound risks related to privacy, bias, discrimination, and human rights violations, particularly when deployed in sensitive domains including law enforcement, immigration, or administering public benefits.

BCGEU, as a long-term responsible investor, understands that a sound approach to AI governance rooted in internationally recognized human rights standards is increasingly material to a company’s ability to manage risk, maintain public trust, and create long-term value. 

Our shareholder proposal aligns with a growing body of international best practice, including the UN B-Tech Project, the OECD Guidelines for Multinational Enterprises, and BSR, who all emphasize the need to root AI governance in a human rights-based approach. According to the UN’s Governing AI for Humanity Report:

To mitigate the risks and harms of AI… human rights must be at the centre of AI governance, ensuring rights-based accountability across jurisdictions.

AI compounds the issues of the already problematic data uses

Thomson Reuters’ products have been directly linked to human rights concerns, particularly through their use by U.S. government agencies including Immigration and Customs Enforcement (ICE) for immigration enforcement, raising serious questions about surveillance, due process, and the rights of migrants and asylum seekers. Further, the Company’s CLEAR investigative software product has enabled law enforcement to access vast quantities of personal information without needing a warrant, even where states have created protections against data sharing with ICE.

Thomson Reuters is currently positioning itself at the forefront of the global AI boom, and its 2025 priorities include evolving transformative AI capabilities through ongoing product innovation and enhancements to flagship products.  The Company is also investing in generative artificial intelligence (genAI), calling its potential “an exponential leap in power”, a characterization that illustrates the scale of the tools’ potential impact. One such offering, Risk Analysis Summary (RAS), integrates with the Company’s CLEAR software. It is marketed as a tool capable of supporting and speeding law enforcement investigations by summarizing and extracting insights from massive datasets like those within CLEAR.

For a company already operating in high-risk data environments, the further deployment of AI and genAI may escalate the complexity and materiality of existing risks, making these concerns more urgent.

Alignment with Global Expectations

Thomson Reuters discloses AI governance measures it has adopted, including a set of Data and AI Ethics Principles. The Company states that these principles were informed by various global standards, including the UNGPs. However, these principles do not explicitly reference the UNGPs or provide detailed insight into how human rights considerations are applied. Proxy advisory firm Glass Lewis, who has recommended shareholders vote in favour of our proposal concurs with this point. They write:

…we note that the Company makes no reference to the UNGPs or other frameworks in the principles…. Upon review, we find these principles to be broad and unspecific in a manner that does not allow shareholders to understand how the Company’s framework is being applied or how human rights considerations are incorporated in their execution.

A Constructive Step Forward

In response to earlier engagement with BCGEU, Thomson Reuters aligned its broader human rights due diligence framework with the UNGPs. We believe that explicitly aligning with the UNGPs offers a clear way for the Company to strengthen its AI governance and create long-term business success.

We have encouraged shareholders to support this proposal as a prudent step toward ensuring that Thomson Reuters continues to lead responsibly in an era of rapidly advancing AI technologies, to protect long-term value, mitigate risk, and uphold human rights. The BCGEU will present its shareholder proposal at Thomson Reuters annual general meeting on June 4 in Toronto.

Read our Investor Brief here.