SHARE’s recent engagement with a U.S.-based health-care company about its use of artificial intelligence illustrates the role investors can play in shaping how this revolutionary technology is developed, adopted, implemented and monitored.
SHARE reached an agreement with United Health Group (NYSE: UNH) to withdraw our proposal asking the company to report on the use and oversight of Artificial Intelligence (“AI”) in its business operations. United Health has agreed to produce a report in the next quarter that outlines the ethical guidelines, governance structures, guardrails, limits, complaints and review mechanisms that surround its use of AI.
The United Health engagement is one of many we expect to support in the coming months and years, following our endorsement of the World Benchmarking Alliance’s 2024 Investor Statement on Ethical AI and our participation in the alliance’s Collective Impact Coalition.
The WBA notes that only 44 out of the 200 digital technology companies it rated in its 2023 Digital Inclusion Benchmark disclosed the principles they follow in the development, deployment and/or procurement of AI tools. That rate is an improvement over previous years, but still a worrisome number given the rapid implementation of AI across so many industries and the level of risk it presents to companies, workers, customers and investors
Investor engagement on AI use
Despite the uncertainty and the rapid pace of change, investors can’t shy away from engaging with investee companies about the many potential pitfalls that might arise from the growth of AI, and to help define guardrails around its use. For example, in our engagements SHARE has focused on:
- Developing safeguards against “algorithmic bias,” a phenomenon in which, through poor design or data choices, a machine-learning system or algorithm reinforces social and economic inequities.
- Oversight of predictive analytics systems, in which data is used to propose answers to questions around health-care coverage, to ensure there are no automated denials of coverage. Even when a company does include a “human in the loop” to oversee decisions, we want to be sure that the company isn’t biasing employees towards acceptance of AI-generated results through performance expectations that weaken their oversight role.
- And, in the case of technology companies like Alphabet (which owns platforms like YouTube and Google), we’re focused on how algorithmic ad targeting, if unchecked, can exacerbate systemic discrimination and human rights violations.
We’ve recently filed a shareholder proposal at Alphabet (NASDAQ: GOOG), asking the company to subject its targeted advertising algorithms to a human rights impact assessment. Absent a settlement within the coming months, the proposal will come up for a vote at the company’s next annual general meeting.
The AFL-CIO Equity Index Funds recently successfully withdrew AI proposals at both Disney (NYSE:DIS) and Comcast (NASDAQ: CMCSA) asking for AI transparency reports that address “whether they have adopted any ethical guidelines to protect workers, customers and the public from harms related to the use of AI.” The AFL-CIO’s proposal at Apple Inc. (NASDAQ: AAPL) went to a vote in February and received 37.5% support from shareholders, leading the CEO to announce improved disclosure later this year.
Setting the stage for further action
The Alphabet vote will help set the bar for AI oversight at a critical player in the global tech field, while the upcoming disclosures from United Health Group, Disney, Comcast and Apple will help SHARE and other investors understand how corporations are developing their AI use and establishing appropriate safeguards.
All are important first steps towards broader actions.
As global standards and our own understanding evolve, investors can engage further to fill in gaps in policy or practice, and establish norms that protect workers, communities, consumers and shareholders from negative consequences, while still harnessing the potential positive effects of this remarkable new technology.
