Harvard Adopts New Shareholder Voting Guidelines on Human Rights, Labor, Data Privacy

Harvard adopted four new shareholder voting guidelines on human rights, labor rights, living wages, and data privacy, according to its most recent annual report.

The guidelines, approved in April 2025, bring Harvard’s total to 31 and provide a framework for how the University and external investment managers at Harvard Management Company approach shareholder votes. HMC oversees Harvard’s $56.9 billion endowment.

Harvard also weighed 20 shareholder resolutions during the 2024-25 academic year at Apple, Meta, Alphabet, and Nvidia on topics including artificial intelligence, climate and emissions, child safety, and data privacy. It supported six proposals, opposed 10, and abstained on four.

The votes offer a window into how Harvard applies its broader guidelines, with the University at times opposing proposals that touched on issues covered by its own policies when committee members found them unclear, duplicative, overly prescriptive, or already addressed.

Harvard maintains two committees focused on shareholder responsibility. The Advisory Committee on Shareholder Responsibility, composed of faculty, students and alumni, develops voting guidelines and recommends how Harvard should vote. The Corporation Commitee on Shareholder Responsibility, composed of members of the Harvard Corporation, reviews the proposed guidelines and makes final voting decisions.

One of the new guidelines states Harvard will generally support resolutions asking a company to commission a third-party report on human rights risks in conflict-affected areas or assess how it is mitigating those risks.

But Harvard opposed a resolution asking Alphabet — the parent company of Google — for a third-party report on whether customers’ use of its products for surveillance, censorship, or military purposes contributes to human rights harm in conflict zones.

The advisory committee concluded that Alphabet was already working to “meet the spirit” of the resolution and that the proposal did not meet the threshold laid out in Harvard’s guideline.

Another new guideline supports third-party assessments of companies’ compliance with commitments to collective bargaining and to not interfere with labor organizing.

A third, focused on living wages, generally supports proposals seeking assessments of companies’ progress toward paying living wages or asking companies to “use their own discretion to set wage policies aimed at meeting a family’s basic needs.”

The final guideline supports proposals seeking greater disclosure of companies’ data privacy policies and assessments of how their data practices could pose privacy or civil rights risks, including whether technology or data are sold to governments.

The guidelines stop short of supporting proposals that mandate a company’s actions. Harvard generally opposes resolutions that directly dictate how companies should manage wages or data privacy.

Most of the resolutions considered were related to data privacy and technology.

Three resolutions — one each at Alphabet, Apple, and Meta — asked the companies to assess risks from improper or unethical use of external data in training artificial intelligence. Harvard supported the proposals at Meta and Alphabet, but abstained on Apple’s.

Some advisory committee members said Apple had already satisfied the proposal’s reporting requirements and that an additional report could be “obsolete” by the time it was completed.

Harvard voted in favor of six resolutions, including one asking Meta to disclose a plan that would achieve emissions reductions to make up for the energy demand from expanding data centers.

But Harvard opposed ten resolutions, including one asking Alphabet to assess the civil rights impact of generative AI and its potential bias against religious and political views. Advisory committee members raised concerns about the breadth of the proposal and its supporting statement, which contained information that members “did not feel aligned with Harvard values.”

Harvard has focused on technology guidelines in the past — the committee established guidance in 2022 on cybersecurity and climate commitments. Harvard also established guidelines on digital media content in 2018.

Harvard abstained from a proposal asking Meta for transparency on data collection, consent mechanisms, and ad targeting. It opposed a similar resolution for Alphabet asking the company to conduct a third-party human rights assessment of AI-driven targeted advertising, citing the proposal’s “lack of clarity.”

Harvard also abstained from a proposal asking Meta to report on combating antisemitism, anti-LGBTQ+, and anti-disability hate after advisory committee members raised concerns about naming three specific groups and excluding others.

“Committee members expressed support for combatting hate in all forms, but there was concern that supporting a proposal that identified only a few groups and not others could cause a misunderstanding,” the report read.

—Staff writer Megan L. Blonigen can be reached at [email protected] and on Signal at mblon.50. Follow her on X at @MeganBlonigen.