Pension Funds Push for Stronger AI Governance, Eyeing Proxy Votes to Curb Growing Risk

Large pension funds and other institutional investors are sharpening their focus on how companies oversee artificial intelligence, treating AI governance as a growing financial risk that can affect long-term portfolio value.

The shift is showing up in the way major shareholders plan to use their influence at annual meetings. Rather than limiting engagement to private discussions with executives, investors are increasingly signaling that they may rely on proxy voting—one of the most direct tools available to shareholders—to push companies toward stronger oversight. That includes the possibility of voting against the re-election of directors if investors judge that boards are not adequately managing AI-related risks and responsibilities.

One of the clearest signs of the trend comes from the California Public Employees’ Retirement System (CalPERS), among the world’s largest public pension funds. CalPERS has, for the first time, incorporated AI-related provisions into its latest proxy voting guidelines. The move effectively places AI alongside other board-level issues that the fund believes can influence corporate performance and shareholder value, and it gives CalPERS a framework for how it will evaluate company behavior when casting votes.

The investor push reflects a growing belief that AI is no longer a niche technology topic but a core governance issue—touching strategy, risk management, compliance and reputational exposure. Institutional shareholders worry that weak oversight can lead to financial harm, whether through flawed deployments, legal and regulatory problems, or broader damage to a company’s brand and trust among customers and partners.

Pension funds and asset managers are responding by demanding clearer accountability for AI decisions, including how companies set internal controls, monitor outcomes, and ensure that responsibility sits at the appropriate level of leadership. For investors who hold diversified portfolios and typically invest for the long term, the concern is not only whether a company can capture the upside of AI, but also whether it can avoid missteps that could erode value.

The development also underscores a broader evolution in shareholder expectations. Governance is increasingly being treated as a forward-looking discipline: investors want boards to demonstrate that they understand emerging technologies and can manage them with the same rigor applied to financial reporting or cybersecurity. Proxy voting—especially votes related to directors—offers investors a concrete mechanism to express dissatisfaction when they believe oversight falls short.

Why it matters

For companies, the message from major shareholders is that AI oversight is becoming a standard component of board accountability, not an optional policy add-on. For retirees and other beneficiaries of large funds, the goal is to reduce the chance that avoidable AI failures translate into lost portfolio value. The next key signal to watch is whether more investors adopt similar voting guidelines—and whether boards respond by formalizing AI governance to preempt challenges at the ballot box.