World Benchmarking Alliance: Ethical AI Collective Impact Coalition

The approach, initiative or process 

The Ethical AI Collective Impact Coalition is an engagement initiative convened by the World Benchmarking Alliance advocating for digital technology companies to implement policies and mechanisms to ensure the ethical development and application of AI, guided by respect for human rights and the principle of leaving no one behind.

Rationale

The proliferation of AI applications in recent years across domains such as finance, health, media and entertainment, advertising, law enforcement and human capital management, has highlighted the underlying risks and opportunities that AI presents, not least regarding the observance of human rights.

 AI can help to improve the accuracy of medical diagnoses, broaden financial inclusion, slow the spread of abusive content online and bolster accessibility for people with disabilities. At the same time, evidence shows that AI may increase the risk of human rights harms such as bias and discrimination, invasion of privacy, denial of individual rights, and other non-

transparent, unsafe and unjust outcomes. If companies do not formulate and implement rights-respecting principles to guide their development, deployment and procurement of ethical AI, they face reputational risk and revenue losses, and society as a whole faces risks of a new wave of systemic problems exacerbated by rapid technology deployment. Not only should companies set up high-level ethical AI frameworks, they should also provide a public account of how they live up to them.

Scope and objectives

The Ethical AI CIC builds on the findings of WBA’s Digital Inclusion Benchmark, which throughout its three iterations (2020, 2021 and 2023), has revealed large transparency gaps in digital companies’ policies and practices on ethical AI. The benchmark covers 200 of the world’s most influential companies operating in technology sectors globally, including emerging markets.

The Ethical AI CIC has three overarching objectives:

1. to raise awareness on the importance of responsible and ethical AI;
2. to increase understanding of the state of play and leading best practices;
3. to improve digital technology companies’ commitments to ethical AI.

 In relation to the last objective, investors and CIC members specifically ask companies to implement, demonstrate and publicly disclose:

• policies or commitments to respect human rights in the company’s development, deployment, and/or procurement of AI tools;
• strong AI governance and oversight across the value chain of AI development and use, including board oversight and senior management responsibility;
• implementation and operationalisation of such principles in the development and use of AI-powered products and services, including identification of salient risks, human rights impact assessments (HRIA), stakeholder engagement, access to remedy and grievance mechanisms for users and anyone impacted by technology use, among others.

Evolving to match the pace of technological change

Since its inception, the initiative has had to develop and mature rapidly to match the pace of AI’s evolution. It has progressed from engaging and encouraging corporates to drive the adoption of ethical AI principles, to a platform that combines collective engagement on AI practices, coupled with investor education that is strongly informed by external expertise facilitated by organisations such as Access Now and BSR.

Phase I (2022–24)

The initiative was launched in September 2022 as the CIC for Digital Inclusion, with a group of 32 investors representing US$6.9 trillion in assets under management (AUM) and 12 civil society organisations. Phase I’s collective investor engagement focused on 44 companies across the US, Europe, Asia and emerging markets, including China, with the goal of encouraging companies to publicly disclose their ethical AI principles and policies. The investor component of the initiative is anchored in the Investor Statement on Ethical AI published in April 2022. That phase ended in September 2023 with the publication of the 2023 Progress Report. During Phase I, 19 of the 44 companies targeted in the collective engagement published their inaugural AI principles.

Phase II (2024–present):

The current phase of the engagement was launched in February 2024, with the CIC growing to 64 investors representing US$8.7 trillion in AUM, expanding engagement efforts from 44 companies in Phase I to 80 companies in Phase II, and deepening the scope of engagement. With companies increasingly introducing AI policies, the engagement needed to go beyond commitments to also encompass implementation and operationalisation of a human rights-based approach to AI use, development and deployment. To date, 37 companies have responded to outreach letters and 26 active dialogues are taking place.

An updated Investor Statement on Ethical AI was published in February 2024 outlining the investor expectations of digital companies. WBA data published in September 2024 shows that 71 companies – a third of those evaluated in the benchmark – had a publicly available set of AI principles, up from 33 in 2021 and 52 in September 2023. Several of these companies have been evaluated across three editions of the Digital Benchmark since 2020, which underscores the value of benchmarking and collective engagement.

As part of Phase II, a new coalition membership status was implemented for Observer organisations to ensure adequate coordination and collaboration with other investor initiatives such as the Investor Alliance on Human Rights (IAHR) and the Emerging Markets Investor Alliance (EMIA). Phase II also welcomed the incorporation of Candriam’s Initiative on Facial Recognition Technology. Candriam, and subsequently Amundi, joined Boston Common Asset Management and Fidelity International as Steering Committee members of the Ethical AI CIC.

Further, Phase II includes continued outreach and expansion to increase the awareness and action of investors and other stakeholders on this critical topic. As part of this effort, the steering committee has set up sector-specific groups dedicated to ethical AI challenges relevant to semiconductors, telecoms and social media companies. The role of these groups is to provide an interactive space for capacity-building and capturing best practices for particular sectors so that investors can leverage them in their company dialogues. Regular monthly calls and education sessions provide learning opportunities for investors and other stakeholders on topics such as building expectations for AI human rights impact assessments and risks involved in generative AI development. The CIC also serves as a platform for sharing updates on relevant policy initiatives and policy events.

Measures to ensure transparency generate outcomes

The initiative offers multiple modalities of communication for stakeholders.

Member communication

Quarterly calls are held for investor members to hear WBA updates, raise questions on related topics and share insights on progress and challenges in company engagement. Investors are asked to submit engagement updates on a regular basis to the WBA team, and are also encouraged to discuss updates on the calls. Investor members have access to a platform that shows engagement progress for each company.

Public accountability

The Ethical AI CIC builds on the findings of WBA’s Digital Inclusion Benchmark, an open-source public good enabling all CIC members, and any other stakeholder including companies, to access company assessments, industry insights, datasets, detailed methodologies and scoring guidelines. The initiative published its first progress report in September 2023, which includes detailed insights, including overarching trends observed in the collective investor-company engagement to ensure that not only CIC members but also other investors and wider stakeholders, are better informed. This report

was accompanied by an online, public session to ensure a global audience could learn from the CIC findings. A second progress report covering Phase II is expected for publication at the end of October 2025, which will be accompanied by an open online launch. The progress reports can be used by other investors for direct engagements or company assessments, multiplying the impact of the CIC.

 A multi-stakeholder approach

A key feature of this initiative is its multi-stakeholder approach, whereby engagement and knowledge-sharing efforts are informed by civil society research. The investor arm is co-led by Boston Common Asset Management, Fidelity International, Candriam and Amundi, while the civil society arm is co-led by Women at the Table and Paradigm Initiative to represent non-governmental organisations (NGOs), research institutions, think tanks and business consulting groups. The group has engaged in multiple events advocating for ethical AI, including:

• a session at UNGA Summit of the Future;
• an open side-event at the 2025 Paris AI Action Summit;
• the publication of a Joint Statement on the Responsible Design, Development and Use of AI.

This group also made a joint submission to the UN Global Digital Compact on digital inclusion topics, including AI regulation, following a series of consultations. This multi-stakeholder systemic stewardship approach is designed to set expectations for and support companies, empower all actors to understand and take action on the topic, and feed in practical experience and expertise to influence the global policy environment at a pivotal moment in the development of this technology.

WBA disclaimer: The World Benchmarking Alliance (WBA) collects publicly available, non-sensitive information in order to develop benchmarks to improve assessments of performance against sustainability goals. WBA is committed to compliance with all applicable laws and regulations in the jurisdictions where it operates.

All information relied on by WBA in the course of its activities — including preparing benchmarks, reports, or other materials (the “Content”) — is publicly available or voluntarily provided by participating companies. All information herein is provided “as-is” with no guarantee of completeness, accuracy or timeliness. WBA will not ask for, and you should not provide or share with WBA, or any other company at a WBA event, any non-public competitively or commercially sensitive information. Such information could include current or forward-looking information about pricing, customers, key contract terms, or strategic plans.

Participants in WBA initiatives, including Allies, companies subject to benchmarking, and any other participants, at all times retain sole discretion over whether and how to act on the information WBA provides. WBA does not act or speak on behalf of its Allies nor any other organisation. No party may claim to represent other Participants at a WBA Initiative or the WBA Alliance without their express and written consent.

All WBA Content and activities are for informational purposes only. WBA does not provide, and you should not construe the Content or activities, as investment, financial tax, legal or other advice. WBA is not a proxy advisory firm and does not provide voting recommendations or other advice on management of an investment portfolio. WBA is not an investor and does not acquire or hold the securities of any company with the purpose or effect of changing or influencing control.  Participants to the WBA Alliance and WBA Initiatives are independent organisations responsible for their own commercial, investment and voting decisions.

PRI disclaimer:This case study aims to contribute to the debate around topical responsible investment issues. It should not be construed as advice, nor relied upon. It is written by a guest contributor. Authors write in their individual capacity – posts do not necessarily represent a PRI view. The inclusion of examples or case studies does not constitute an endorsement by PRI Association or PRI signatories.