Thomson Reuters corp. | Amend AI governance framework to align with UNGPs at Thomson Reuters corp.

Status
Filed
AGM date
Previous AGM date
Resolution details
Lead filer
B.C. General Employees' Union (BCGEU)
Resolution ask
Adopt or amend a policy
ESG theme
  • Social
ESG sub-theme
  • Digital rights
Type of vote
Shareholder proposal
Filer type
Shareholder
Company sector
Consumer Discretionary
Company HQ country
Canada
Resolved clause
Shareholders request that the Board of Directors of Thomson Reuters Corporation amend its artificial intelligence (AI) governance framework to be consistent with its overall approach to human rights due diligence, in alignment with the UN Guiding Principles on Business and Human Rights (UNGPs). The Company should assess the suitability of its Trust Principles for governing AI-related risks and disclose whether additional measures are needed to ensure responsible AI use and development.
Supporting statement
Thomson Reuters has made significant investments in artificial intelligence (AI) and claims to be at the forefront of generative AI (GenAI), offering Risk Analysis Summary (RAS) which works within its Consolidated Lead Evaluation and Reporting (CLEAR) investigative software. The Company’s approach to AI governance is informed by its Data and AI Ethics Principles. According to the first principle, TRI’s use of data and AI is informed by the Trust Principles. The Trust Principles, designed for news integrity and independence, are insufficiently robust to address AI-related human rights risks. In response to previous engagement from shareholders, TRI acknowledged the unsuitability of the Trust Principles to govern human rights risk and disclosed alignment with the UNGPs instead. Thomson Reuters leases its CLEAR software to police forces and to U.S. government agencies including Immigration and Customs Enforcement (ICE) where violations of human and privacy rights have been reported. AI systems, particularly those used in legal and investigative products such as CLEAR, raise significant concerns related to privacy, due process, equal protection, and human rights. AI-driven risk assessment and predictive analytics systems have been criticized for potential bias and discriminatory outcomes.12

The UNGPs, OECD Due Diligence Guidelines, and guidance from the UN B-Tech Project recommend incorporating a human rights-based approach to AI governance. This entails:
• Conducting ongoing human rights due diligence in AI development and deployment;
• Disclosing governance structures, impact assessments, and risk mitigation measures;
• Consulting with affected stakeholders, including civil society, marginalized groups, and employees;
• Conducting human rights impact assessments for AI applications, particularly in high-risk areas;
• Providing clear processes for addressing grievances related to AI-related harms; and
• Ensuring alliance with global standards and compliance with evolving regulatory frameworks.

Peer companies such as Microsoft have implemented robust Responsible AI Standards that go beyond broad ethicalcommitments, incorporating clear governance mechanisms and oversight structures. Companies such as IBM, OpenText, and TELUS have signed on to Canada’s Voluntary Code of Conduct on Advanced Generative AI, which includes post-deployment monitoring of AI systems for harmful impacts.3 Thomson Reuters is embracing the future potential of AI and GenAI, creating opportunities for long term value creation. AI developments can enhance efficiency but must be responsibly managed to ensure effective risk management. Thomson Reuters has already adopted robust human rights governance and has an opportunity to extend this same level of governance to its AI offerings. Transparency, oversight, and accountability in AI governance is critical for the company to ensure the protection of fundamental rights.

DISCLAIMER: By including a shareholder resolution or management proposal in this database, neither the PRI nor the sponsor of the resolution or proposal is seeking authority to act as proxy for any shareholder; shareholders should vote their proxies in accordance with their own policies and requirements.

Any voting recommendations set forth in the descriptions of the resolutions and management proposals included in this database are made by the sponsors of those resolutions and proposals, and do not represent the views of the PRI.

Information on the shareholder resolutions, management proposals and votes in this database have been obtained from sources that are believed to be reliable, but the PRI does not represent that it is accurate, complete, or up-to-date, including information relating to resolutions and management proposals, other signatories’ vote pre-declarations (including voting rationales), or the current status of a resolution or proposal. You should consult companies’ proxy statements for complete information on all matters to be voted on at a meeting.