AI and Proxy Voting: How AI Is Disrupting Corporate Governance Advisory Services

On January 7, 2026, JPMorgan Chase announced that its asset management division would stop using external proxy advisory firms for U.S. voting decisions and replace them with an internally developed AI platform called Proxy IQ. The announcement was a watershed moment for institutional governance. JPMorgan Asset & Wealth Management manages trillions of dollars in assets and holds significant voting stakes across more than 3,000 publicly listed U.S. companies. Its decision to abandon ISS and Glass Lewis — the two firms that together control more than 90 percent of the proxy advisory market — and replace them with an AI system was simultaneously a sharp institutional rebuke of the proxy advisory industry and the first large-scale deployment of AI as a direct substitute for human governance advisory services.

This post examines what Proxy IQ is designed to do, the fiduciary and regulatory framework within which it operates, the structural critique of proxy advisory firms that JPMorgan’s move reflects, and what this shift means for companies managing shareholder engagement and corporate governance.

What Proxy IQ Is — and What It Replaces

Proxy advisory firms like Institutional Shareholder Services (ISS) and Glass Lewis emerged to solve a real problem. Large institutional investors — pension funds, mutual funds, sovereign wealth funds, and asset managers — hold shares in hundreds or thousands of publicly traded companies. Each of those companies holds an annual shareholder meeting at which investors are asked to vote on a range of matters: electing directors, approving executive compensation plans, ratifying auditors, and acting on shareholder proposals. For a large institutional investor, developing and applying independent voting positions across thousands of proposals from thousands of companies every proxy season is an enormous undertaking.

Proxy advisory firms built businesses that effectively outsourced this work. They developed voting guidelines — essentially default policies for how institutional investors should vote on standard categories of proposals — and applied those guidelines at scale across their client bases. By many estimates, their recommendations move institutional voting in a substantial majority of contested situations. A company that receives an “against” recommendation from ISS on its executive compensation program faces a materially increased likelihood of a failed say-on-pay vote.

The power and the criticism of proxy advisors stem from the same source: they apply standardized guidelines at massive scale. Critics, including many corporate issuers, argue that the guidelines are too rigid, often reflect the policy preferences of the advisory firm’s analysts rather than the specific interests of the institutional investor’s clients, and create conflicts of interest where the advisory firm also sells governance consulting services to the companies it rates. JPMorgan CEO Jamie Dimon was characteristically blunt, publicly calling the proxy advisory firms “incompetent” in the context of announcing the Proxy IQ transition.

Proxy IQ replaces this external advisory relationship with an internal AI system. The tool runs on JPMorgan’s Spectrum investment data platform and is designed to gather and analyze data from company filings, corporate governance disclosures, compensation data, and other sources across the full universe of U.S. companies where JPMorgan Asset Management holds shares. The system applies JPMorgan’s own voting guidelines rather than ISS or Glass Lewis guidelines, generating voting recommendations at the scale and speed necessary to cover 3,000-plus annual meetings per proxy season.

The Fiduciary Framework

Proxy voting by institutional investors is not a discretionary act — it is a fiduciary obligation. Investment advisers registered with the SEC owe their clients a duty to vote proxies in the clients’ best interests. The SEC has reinforced this position over multiple administrations. In the current regulatory environment, the SEC has directed investment advisers to re-examine longstanding practices of routinely delegating voting decisions to third-party proxy advisors and to ensure that voting decisions are grounded in each adviser’s own analysis of what serves client interests.

JPMorgan’s public rationale for the Proxy IQ transition tracks this regulatory posture precisely. The firm stated that the change reflects its goal of exercising proxy voting authority based solely on what it determines to be in the best interests of its clients — not in the interests of a third-party service provider whose guidelines may diverge from the investment manager’s own views. The SEC’s informal engagement on the topic has signaled that proxy voting is itself a fiduciary function and that investment advisers should not default to third-party recommendations as a substitute for their own fiduciary analysis.

This creates an interesting governance question that Proxy IQ is designed to answer: how does a large institutional investor exercise genuine independent fiduciary judgment over thousands of companies across a compressed proxy season, without the operational infrastructure of a large proxy advisory organization? JPMorgan’s answer is to build that infrastructure in-house as an AI system — one that applies the firm’s own guidelines, on the firm’s own data, without the conflicts of interest and standardization tradeoffs inherent in outsourcing the function.

Whether an AI-generated recommendation constitutes “independent fiduciary judgment” in the regulatory sense remains an open question. The SEC’s guidance emphasizes that AI tools could serve as a useful aid, provided that investment advisers retain accountability and make accurate disclosures about their use. The tool is a means of analysis, not a substitute for accountability. JPMorgan’s structure — where Proxy IQ generates recommendations that human governance professionals at the firm apply — is designed to preserve the human accountability layer while using AI for the analytical scale.

Implications for ISS and Glass Lewis

The structural consequences for the proxy advisory industry are significant. ISS and Glass Lewis have operated as de facto standard-setters for institutional governance voting for decades. Their combined market share — estimated at over 90 percent of the proxy advisory market — reflects the network effects of the business: the more institutional investors use the same guidelines, the more companies calibrate their governance practices to satisfy those guidelines, which further entrenches the advisory firms’ position.

JPMorgan’s move disrupts this dynamic by signaling that major institutional investors can — and may — develop independent governance views at scale. The combination of internal AI infrastructure and a willingness to diverge from industry-standard guidelines creates a governance landscape where companies cannot simply assume that a “passing” ISS or Glass Lewis rating will translate into institutional investor support across the board.

For governance advisors counseling companies through proxy season, the lesson is that institutional voter analysis can no longer rely exclusively on understanding ISS and Glass Lewis guidelines. As more institutional investors develop independent AI-powered governance platforms, each with its own data inputs and voting guidelines, the institutional voting landscape will become more heterogeneous and less predictable from a single-guideline baseline.

It is also worth noting the regulatory environment in which JPMorgan’s announcement arrived. The current SEC and the current Congress have been more skeptical of proxy advisory firms than their predecessors, particularly with respect to the advisory firms’ governance guidelines on topics like board diversity and environmental, social, and governance (ESG) proposals. An executive order from the current administration reinforced this skepticism. JPMorgan’s move to replace external advisory firms with an internal AI system simultaneously sidesteps this political controversy — Proxy IQ applies JPMorgan’s own guidelines, not the contested policy frameworks of ISS or Glass Lewis.

What This Means for Companies Managing Shareholder Engagement

For companies — particularly public companies and those approaching an IPO — JPMorgan’s Proxy IQ transition carries several practical governance implications.

Shareholder engagement becomes more important, not less. When institutional voting decisions are generated by AI systems applying each investor’s proprietary guidelines, the margin for error in shareholder engagement narrows. Companies cannot rely on satisfying a single, publicly available ISS or Glass Lewis framework and assume institutional support will follow. Direct engagement with major institutional shareholders — explaining governance decisions, compensation design, board composition rationale, and strategic context — becomes the principal mechanism for ensuring that the institutional investor’s AI system has the information and context necessary to reach a well-informed vote.

Disclosure quality directly affects AI-powered analysis. AI governance systems like Proxy IQ necessarily work from the data that companies disclose in their proxy statements, sustainability reports, and governance disclosures. If a company’s disclosures are sparse, poorly organized, or fail to address the factors that large institutional investors weigh in their governance frameworks, the AI system will fill gaps with default assumptions or comparisons to peer companies. Companies that invest in clear, comprehensive, and well-organized governance disclosures are better positioned in an AI-analyzed environment than companies whose disclosures force the AI to make inferences.

Executive compensation design faces new scrutiny. Say-on-pay votes — where shareholders vote on executive compensation programs — are among the most consequential annual meeting items for company boards. Traditional proxy advisory guidelines on compensation have been relatively predictable, if contested. As institutional investors apply their own AI-powered analysis to compensation data, the inputs that matter most will be those that align with each investor’s own governance philosophy. For companies with majority-institutional ownership, understanding the compensation governance frameworks of their largest shareholders — rather than focusing exclusively on ISS guidelines — becomes an increasingly important part of the compensation committee’s annual planning.

Board composition and director qualifications are under real-time analysis. AI systems analyzing governance across thousands of companies can identify patterns in board composition — average tenure, skill set distribution, independence ratios, director commitments — more quickly and consistently than human analysts reviewing individual proxy statements. Boards whose composition lags behind peer norms on factors that major institutional investors prioritize will face more consistent and better-documented negative recommendations than in an era when those analyses were performed by human analysts under time pressure.

Broader Industry Trajectory

JPMorgan is unlikely to remain the only large institutional investor to build AI-powered governance infrastructure. The economics are compelling: an in-house AI system, once built and maintained, replaces recurring advisory fees while producing voting recommendations aligned with the investor’s own policy preferences. For large institutional investors that have already invested heavily in data and analytics infrastructure for investment management, extending that infrastructure to governance decisions is a natural evolution.

The governance consequences for the broader ecosystem will play out over multiple proxy seasons. A world in which the largest institutional investors each apply their own AI-powered governance frameworks — rather than a shared ISS or Glass Lewis baseline — is a more fragmented governance environment. Companies will need to understand the governance priorities of each major institutional shareholder rather than managing to a single industry standard. That understanding requires direct engagement, transparent disclosure, and governance decisions that can be explained and justified, not merely calibrated to pass a third-party checklist.

For smaller companies with less capacity for sophisticated shareholder engagement, this shift may increase governance complexity. For companies with robust investor relations functions and governance programs designed around genuine accountability to shareholders rather than checkbox compliance, it is an opportunity to differentiate.

The deeper governance principle at stake is one that JPMorgan’s move makes explicit: proxy voting is a fiduciary act that belongs to the investor. Delegating it entirely to a third party — whether a human advisory firm or an AI system that operates without genuine oversight — does not satisfy the obligation. What Proxy IQ represents, at its best, is an institutional investor taking that obligation seriously enough to build the infrastructure necessary to discharge it independently, at scale.

This post is for general informational purposes only and does not constitute legal advice. Reading this post does not create an attorney-client relationship. If you have questions about your specific situation, consult a qualified attorney.