How bringing stakeholders together can influence companies and facilitate action

To achieve the SDGs through systemic change, actors of different scales, across diverse contexts and over time, must work together. WBA is actively bringing together key stakeholders to collectively hold companies to account using WBA’s benchmark results and insights to drive progress on catalytic issues.
Collective Impact Coalitions (CICs) provide a space for WBA Allies and stakeholders to bring forward collaborative action based on the data and evidence provided by WBA benchmarks. CICs are multistakeholder and time-bound, bringing together actors ranging from large institutional investors to civil society, academia, and business platforms to coordinate and collaborate on actions to drive positive change from companies on systemically important topics. CICs also provide an appropriate mechanism to initiate action on emerging and catalytic topics (for example, ethical AI) and launch initiatives that can continue to grow and evolve within other host organisations, including WBA Allies.
Building on our experience since the launch of our first CIC (the Social CIC) in 2022, it is worth exploring how additional stakeholder groups (such as multilateral development banks, commercial banks, consultancies, and others) can actively engage with CICs, alongside identifying specific outcomes associated with their participation.
The Collective Impact Coalition for Ethical AI
The CIC for Ethical AI (formerly Digital CIC) was launched in September 2022. This CIC pushes technology companies to advance ethical AI policies and practices. It builds on the findings of the Digital Inclusion Benchmark (DIB), which has revealed large transparency gaps in companies’ disclosures on ethical AI.
In September 2023, we published the first CIC progress report. It outlines the CIC investor engagement process and outcomes, as well as civil
society actions, including a joint submission with recommendations for the United Nations Tech Envoy on AI Governance. Furthermore, the report captures new company performance data on ethical AI. While progress in the adoption of ethical AI principles is still slow, there are clear signs of investor influence over company performance. Most companies are receptive to investor outreach on ethical AI as part of the CIC engagement, though a minority remain unresponsive or even push back.
Out of the 44 companies that investors have attempted to engage with since the inauguration of the CIC, 28 – about two-thirds – have responded
to our collective outreach. Eleven others remained unresponsive or explicitly declined to dialogue. Several argued that the risks of AI were either not material or not applicable to their businesses. Currently, 52 out of the 200 companies included in the benchmark have publicly available AI frameworks. This is up from 44 in the March 2023 DIB findings. The CIC currently has 62 active members – 50 investors and 12 civil society groups. Since the launch of the CIC for Ethical AI, 19 of the 200 companies evaluated in the 2023 DIB have announced their inaugural AI principles. CIC members had engaged with all 19 of these companies throughout the year.
As of September 2023, 52 companies – more than a quarter of those we evaluate – have a publicly available set of principles on AI, up from 33 in the 2021 DIB. Several of these companies have been evaluated across three editions of the benchmark beginning in 2020, underscoring the value of collective engagement.
Our learnings from 2023
Despite success in terms of increased investor action, as well as company disclosures on AI principles, WBA experienced a challenge in bringing together the CIC’s two main stakeholder groups – investors and civil society groups – towards collective action. We plan to draw up a strategic plan to support the work of the CIC’s civil society cohort by focusing on experts and organisations from underrepresented countries and communities.
Our goal is to increase interaction and coordination between CIC investors and other stakeholders and push tech companies to demonstrate stronger commitments to ethical AI principles. This story is an excerpt from our 2023 Annual Report.