Unveiling Key Trends in AI Shareholder Proposals

Overview

AI is the hot topic of 2024. Companies see the opportunity to gain efficiency and drive growth through the adoption of AI, while not ignoring upcoming regulatory obligations and rising investor expectations. Investors want companies to adopt AI to enhance or at least uphold their competitiveness, but they are also increasingly worried about the impact of AI usage on people and the planet. This concern has led to increased scrutiny of AI governance and a rise in AI-related shareholder proposals in the US.

In March 2024, we previewed the landscape for AI proposals and highlighted how proxy advisors approached them at Apple and Microsoft. Now that the 2024 AGM season is largely over, this report provides a comprehensive review based on 23 AI proposals submitted by shareholders at US companies between 1 January 2023 and 30 June 2024 (see the full list of proposals at the end of the report). Here are our key observations:

• Proposal types. The emergence of new types of proposals contributed to a doubling in the number of proposals observed in 2024. The main contributor to this increase was proposals asking for transparency on companies’ AI usage and ethical guidelines, which were submitted at seven companies. For the first time in 2024, shareholders also asked for specific attributions of board responsibilities aimed at improving AI oversight, as well as disclosures related to the social implications of the use of AI on the workforce. Like in 2023, shareholders asked for more disclosure regarding human rights risks as well as misinformation and disinformation risks.
• Filers. Proposals were filed by a wide range of investors and particularly those concerned with the social risks of AI. These included union funds, socially responsible investment funds, faith-based investors and pension funds.
• Targeted industries. Proposals spread across a wider range of companies and industries in 2024. While Big Tech companies (Alphabet, Amazon, Apple, Meta and Microsoft) received most of the proposals filed in 2023 and 2024, smaller media and entertainment companies, a restaurant chain and a healthcare company were targeted for the first time in 2024. This is consistent with AI technologies being increasingly used across all industries and social concerns spreading across sectors as well.
• Company responses. Companies have approached AI proposals differently. Some unsuccessfully attempted to remove the proposal from their AGM ballot through engagement with the US Securities and Exchange Commission (SEC). Certain companies made commitments to improve disclosures resulting in the withdrawal of the proposal. However, in most instances, the proposal was included on the ballot with a recommendation from the board that shareholders vote against the resolution. Supporting their recommendations, boards highlighted the policies and processes already in place, the disclosures already available, as well as the risks associated with the disclosure of strategic information.
• Proxy advisor responses. Proxy advisors have approached AI proposals like ESG shareholder resolutions. As such, they have been inclined to support proposals seeking additional disclosures related to the management of material risks.
• Voting results. Although all proposals failed to garner majority support, some received relatively high levels of support. The proposals that received the highest levels of support were those asking Apple and Netflix to provide transparency reports on their AI usage. These were supported by 37.5% and 43.3% of the votes cast, respectively. The relatively high levels of support observed on a number of proposals confirm that an increasing number of investors want to understand how companies manage the material opportunities and risks associated with the deployment of AI.

The trend is clear shareholder proposals focused on AI are increasing and are garnering support. While shareholder proposals as an escalation tactic have generally been more popular in the US than in other regions of the world, companies globally need to be prepared for increasing investor engagement on AI. Investors should be expected to approach AI risks no differently than they would approach other risk mitigation issues: they will want to see that the board and company are thinking critically about vulnerabilities and taking active steps to address them. Investors such as LGIM and Norges have already publicly disclosed their expectations regarding AI oversight.

Proposal types

In 2023 and 2024, shareholders had two main asks: transparency reports(seven proposals in 2023; 14 in 2024) and improved oversight (two proposals in 2024).

Transparency reports can be categorised as follows:

• Report on AI use and ethical guidelines: These proposals asked that companies provide disclosures on their use of AI and that they disclose any ethical guidelines adopted in relation to AI usage. In certain instances, these proposals asked that companies clarify the role of the board in overseeing AI usage. This type of proposal, observed for the first time in 2024, was the most frequently submitted during this AGM season, with seven companies receiving such proposals.
• Report on human rights risks: These proposals called for independent third-party reports on the risks that AI usage poses to human rights, with a particular focus on facial recognition technology, which may impact people’s privacy – and targeted advertising applications, due to their potential to exacerbate discrimination against certain categories of the population.
• Report on misinformation and disinformation risks: Shareholders asked for reports on the risks to the company and its stakeholders of AI-generated misinformation and disinformation. Shareholders also asked that these reports cover the measures taken to mitigate (potential) harms.
• Report on social implications for the workforce: This proposal, filed at Chipotle, asked for a report on the principles the company will use to address and measure the impact on the workforce of advanced technologies, including AI and automation.
• Report on board oversight: A proposal, filed at Amazon, asked for an independent assessment of the Audit and Compliance Committee’s role in ensuring effective oversight of material risks to stakeholders from the company’s operations, including in relation to its use of AI.
• Report on algorithmic systems: This proposal, filed at Alphabet, asked for additional disclosures of quantitative and qualitative information on algorithmic systems, like how they are used to target and deliver advertisements, error rates and the impacts on user speech and experiences.

In 2024, two companies also received proposals directly asking for specific attributions of board responsibilities with the aim of improving AI oversight.

• Create New Board Committee: The proposal asked that Amazon create a new board committee dedicated to addressing the human rights risks associated with the use of AI.
• Amend Charter of Audit and Compliance Committee: This proposal asked that Alphabet amend the charter of its Audit and Compliance Committee to make clear that it is responsible for the oversight of AI activities.

The chart below shows how many proposals of each type were filed in 2023 and 2024, highlighting the growth in the number and types of proposals observed in 2024. However, it is interesting to note that a number of Big Tech companies started receiving AI proposals even before 2023. For example, Amazon was first asked to report on the human rights risks associated with its facial recognition technology in 2019; Meta was asked to report on the impacts of its targeted advertising policies and practices in 2022; and, Alphabet was asked to report on its algorithmic systems in 2022.

Filers

AI proposals have been filed by a variety of investors and advocacy groups, primarily categorised as union funds, socially responsible investment funds or advisors, faith-based investors or pension funds. Some of these stakeholders have a particular interest in mitigating the social risks associated with AI. For example, union funds are particularly concerned about the potential implications of AI usage on the workforce.

Union funds

• The American Federation of Labor and Congress of Industrial Organizations (AFL-CIO) entered the AI space in 2024 and was the most active filer during this AGM season. It submitted proposals at five technology, media and entertainment companies (Apple, Comcast, Disney, Netflix and Warner Bros), requesting reports on AI usage and ethical guidelines. It also filed a different proposal at Amazon, asking for the creation of a new board committee to address the human rights risks associated with AI.
• The International Brotherhood of Teamsters General Fund was one of the few shareholders in 2024 to target a company outside of the technology, media and entertainment industries by asking the restaurant chain Chipotle to report on its plans to address the implications on the workforce from its use of AI and automation.

Socially responsible investment funds or advisors

• Arjuna Capital submitted proposals at three Big Tech companies (Alphabet, Meta and Microsoft) in 2023 and 2024, asking for reports on the risks of AI-generated misinformation and disinformation.
• The Shareholder Association for Research and Education (SHARE) filed proposals at Apple in 2023 and 2024, requesting a report on the human rights risks associated with their targeted advertising practices. In 2024, they also submitted a proposal at UnitedHealth Group asking for additional disclosure on their use and oversight of AI.
• Trillium Asset Management initially submitted proposals in 2022 and 2023 at Alphabet asking for transparency reports on its algorithmic systems. However, in 2024, Trillium escalated its engagement by submitting a proposal requesting the Audit and Compliance Committee’s charter be updated to reflect its responsibility for AI.
• Harrington Investments has submitted the same proposal at Amazon for the last six years, asking for a third-party report on the human rights risks associated with the use of their facial recognition technology. In 2023, the investment manager also filed a proposal at Alphabet, asking for an independent assessment of the Audit and Compliance Committee’s role in ensuring effective oversight of material risks to stakeholders from the company’s operations, including in relation to its use of AI.

Faith-based investors

• The American Baptist Home Mission Society has submitted the same proposal at Amazon for the last five years, requesting an independent study assessing whether customers’ use of its products could enable mass surveillance leading to human rights violations.
• Mercy Investment Services has submitted the same proposal at Meta for the last three years, asking a third-party report on the human rights impacts of its targeted advertising practices.

Pension funds

• The New York City Retirement Systems submitted a proposal in 2024 at Paramount seeking a report on the company’s use of AI, the board’s role in overseeing it and its ethical guidelines.

The chart below describes the number of proposals submitted by each of these filers in 2023 and 2024, highlighting repeated filings by certain investors as well as the leading role of AFL-CIO in 2024.

Targeted industries

In 2023, all proposals were filed at Big Tech companies (Alphabet, Amazon, Meta and Microsoft). In 2024, Big Tech firms remained a key target group (Alphabet, Amazon and Apple) while other media and entertainment companies received proposals for the first time (Comcast, Disney, Netflix, Paramount and Warner Bros). A restaurant chain (Chipotle) and healthcare company (UnitedHealthcare Group) also received proposals for the first time in 2024. The chart below provides a breakdown of AI proposals per industry group (GICS), treating Big Tech (Alphabet, Amazon, Apple, Meta and Microsoft) as a separate category.

Company responses

The companies that received shareholder proposals in 2024 chose to respond in various ways and putting the proposals to a vote at AGMs was not always the initial response.

Unsuccessful attempts to exclude AI proposals from the AGM agenda

Apple and Disney engaged with the SEC to seek to exclude AFL-CIO’s proposal from the agenda of their 2024 AGM. Both companies argued that the proposal, requesting a transparency report on their AI usage, was directly related to their ordinary business operations. Apple also argued that the proposal sought to micromanage the company. However, the SEC denied both requests, concluding that “the proposal transcends ordinary business matters and does not seek to micromanage the company.”

Commitments leading to the withdrawal of AI proposals

According to Responsible Investor, AFL-CIO agreed to withdraw its proposal at Comcast and Disney following commitments from these companies to enhance their disclosures on their use of AI. Similarly, SHARE announced in March 2024 that it agreed to withdraw its proposal at UnitedHealth Group following the company’s commitment to produce a report in the next quarter outlining its “ethical guidelines, governance structures, guardrails, limits, complaints and review mechanisms” surrounding its use of AI. Of note, the company’s latest sustainability report, released in May 2024, included a dedicated section on the “responsible use of artificial intelligence and machine learning”, covering their AI usage, governance structure, principles and guidelines, as well details about an independent assessment of their “responsible use of AI program.”

Board recommendations and supporting rationale

The remaining companies included the AI proposals in their proxy statements with recommendations from their boards that shareholders vote against these resolutions. In support of their recommendations, boards raised concerns regarding the disclosure of strategic information, reiterated their companies’ commitments to human rights, and explained they had sufficient policies and processes in place that addressed shareholder concerns, sometimes specifically emphasising investments in safety and security.

Proxy advisor responses

Major institutional investors typically receive the research and vote recommendations produced by one or more proxy advisors, most often ISS or Glass Lewis (or both), prior to voting at shareholder meetings. For a new and complex topic like AI, the influence of these actors could be even more important than for more standard resolutions.

In an earlier report, we detailed how proxy advisors reviewed a number of factors including risks, regulation, controversies, company and peer disclosures, to formulate their recommendations on AI proposals.

Glass Lewis supported 17 of the 20 AI proposals that went to a vote in 2023 and 2024. Such support is generally consistent with proxy advisors’ overarching approach to assessing ESG shareholder initiatives, where they are inclined to support proposals that seek enhanced disclosures related to material risks. Interestingly, they recommended that shareholders oppose the proposal asking that Amazon create a new board committee dedicated to addressing the human rights risks associated with AI usage, which was could be seen as overly prescriptive. They also recommended that shareholders vote against the proposal asking that Microsoft provide a transparency report detailing the potential risks to the company’s operation and finances and to broader public welfare arising from misinformation and disinformation disseminated or generated through AI. In this instance, they considered that Microsoft had already provided substantial disclosures regarding its considerations related to AI and the measures it had implemented to mitigate associated risks. While we only have information regarding the vote recommendations of ISS for seven proposals, we note that they supported six of these resolutions. Like Glass Lewis, they recommended that shareholders oppose the resolution filed at Microsoft because the company had already provided sufficient disclosure.

Voting results

Although none of the AI proposals that went to a vote were approved by shareholders, some received relatively high levels of support. Notably, the proposal at Netflix garnered 43.3% of votes in favour, falling just 6.7 percentage points short of the majority threshold.

Three proposals were supported by more than 20% of the votes cast in 2024. These high-scoring proposals at Apple, Netflix and Warner Bros were all filed by the AFL-CIO and requested transparency reports on the companies’ AI use and ethical guidelines. Glass Lewis had recommended that shareholders vote in favour of the proposals filed at those companies. The vote results confirm that when the risks related to AI use are perceived as material, a substantial number of institutional shareholders are eager to better understand these risks and the measures taken by companies to manage them.

Proposals receiving support levels between 10% and 20% focused on transparency requests concerning specific AI risks. At Alphabet and Meta, given the companies’ concentrated ownership structures, the medium support levels likely reflect a significant number of institutional investors’ desire for enhanced disclosures. Conversely, given the more dispersed ownership structures at Amazon and Chipotle, the medium support levels reflect a more moderate support from institutional investors regarding the need for enhanced disclosures.

Among companies with the lowest levels of support for AI proposals, two received requests to alter the responsibilities of their board members with regard to AI oversight. Most investors tend to view such proposals as overly prescriptive, likely driving the lower support. The lowest level of support was observed at Paramount, which is unsurprising given that a single shareholder holds 77% of the company’s shares.

We also note that the three proposals withdrawn at Comcast, Disney and UnitedHealth Group were all requests for reports on AI usage, ethical guidelines and board oversight. Given the high levels of support for similar proposals that went to a vote, it is plausible that these companies anticipated shareholder demands and made substantial commitments, leading to the withdrawal of these proposals.

Looking Forward

The growing numbers of proposals, proposal types, filers, targeted companies and industries reflect a clear trend. While AI technology is being adopted across all industries, AI proposals have only recently started to reach outside of the technology, media and entertainment sector. We expect further expansion in 2025. Voting results at Apple and Netflix, as well as the disclosure commitments obtained at Comcast, Disney and UnitedHealth Group may further motivate union funds and other investors to file similar proposals at other companies.

While only one proposal in our dataset – filed at Chipotle – was specifically focused on the impact of AI and automation on the workforce, we anticipate that such concern may play a bigger role in 2025. It is interesting to note that such concerns may not only be due to companies seeking to directly enhance or protect their competitiveness, but they may also arise in connection with companies’ efforts to pursue ESG objectives, such as reducing their carbon footprint. As a result, we have seen two proposals in 2024 (not included in our dataset) requesting transparency reports on climate change strategies, which included company automation initiatives. These two proposals, filed by unions at Ryder System and Republic Services, referenced the International Labour Organisation’s guidelines on a Just Transition, which focus on ensuring that workers in carbon-intensive industries are not “left behind” as renewable technology becomes more widespread, and instead are appropriately upskilled to new, sustainability-friendly positions. While these proposals do not always specifically name AI and instead use the terms advanced technologies or automation to refer to technology like autonomous vehicles, they are still seeking disclosures on AI’s impacts on workers and communities. In our view, companies with large workforces susceptible to the impact from a new major automation strategy should expect engagement from unions and union funds.

AI proposals are only one manifestation of shareholder engagement on AI. Investors such as LGIM and Norges have set out their expectations regarding AI oversight, and other investors have initiated engagement campaigns asking companies to demonstrate how they manage the risks associated with the deployment of AI. Companies should proactively address these demands, adopting a pragmatic approach to AI governance such as the one outlined in our Responsible AI Governance report.

 

The full report can be found here.