JPMorgan Chase & Co. is making a decisive break with long-standing Wall Street practice by abandoning traditional proxy advisory firms and turning instead to artificial intelligence to guide how it votes shares at U.S. company meetings. The move marks a significant shift in corporate governance and underscores how deeply AI is reshaping decision-making in global finance.
Beginning with the current proxy season, JPMorgan’s asset management arm will no longer rely on external proxy advisors for recommendations on shareholder votes. For decades, such firms have played a central role in advising institutional investors on how to vote on issues ranging from board appointments and executive compensation to mergers, shareholder proposals, and environmental and social policies. JPMorgan’s decision signals growing dissatisfaction with that model and growing confidence in AI-driven analysis.
At the heart of the shift is an internally developed AI system designed to analyse proxy statements, shareholder proposals, and governance data at scale. The system is intended to evaluate thousands of resolutions across U.S. companies, apply JPMorgan’s internal voting policies, and generate recommendations that portfolio managers can use when casting votes on behalf of clients. Executives at the bank describe the initiative as a way to bring greater independence, consistency, and efficiency to the voting process.
Proxy advisors have long been criticised for their outsized influence on corporate governance. Because many large investors follow their recommendations, a single advisory firm’s guidance can significantly shape voting outcomes. Corporate leaders and some policymakers have argued that this concentration of influence can lead to one-size-fits-all approaches that do not always reflect company-specific circumstances or investor preferences. JPMorgan’s move appears aimed at reducing that dependency.
The shift also reflects broader changes in how asset managers view their fiduciary responsibilities. Rather than outsourcing key judgments, JPMorgan is seeking to internalise analysis and align voting decisions more closely with its own research and the priorities of its clients. By using AI, the firm believes it can process more information more quickly while tailoring decisions to nuanced policy frameworks.
Technology has become central to JPMorgan’s strategy in recent years. The bank invests billions of dollars annually in technology and employs thousands of engineers, data scientists, and AI specialists. AI tools are already used across its operations, from fraud detection and risk management to trading and customer service. Applying similar technology to proxy voting is seen internally as a logical next step.
Still, the move raises important questions about transparency and accountability. Proxy advisors traditionally provide detailed explanations for their recommendations, which investors and companies can scrutinise. An AI-driven system, by contrast, relies on complex models that may be difficult for outsiders to fully understand. Governance experts warn that investors will need clarity on how such systems are trained, what data they prioritise, and how potential biases are addressed.
JPMorgan has said that AI will not replace human judgment entirely. Portfolio managers and governance specialists will retain oversight and make final decisions, using AI outputs as a tool rather than a mandate. The bank argues that this hybrid approach combines the speed and scale of AI with human expertise and ethical responsibility.
The implications of JPMorgan’s decision extend beyond the firm itself. As one of the world’s largest asset managers, its practices are closely watched by peers. If the AI-based approach proves effective and withstands regulatory scrutiny, other large investors may consider developing similar systems. That could gradually erode the dominance of proxy advisory firms, reshaping the governance ecosystem.
At the same time, smaller asset managers may struggle to follow suit. Developing and maintaining sophisticated AI systems requires significant investment and technical expertise. For many firms, proxy advisors may remain the most practical option. This could lead to a more fragmented landscape, with large institutions relying on proprietary AI while smaller players continue to depend on external guidance.
The move also comes at a sensitive moment in U.S. corporate governance. Shareholder votes have become increasingly contentious, particularly around issues such as climate strategy, diversity policies, and executive pay. How JPMorgan’s AI system approaches these topics will be closely examined by companies, activists, and regulators alike.
Corporate boards are already preparing for the possibility that voting patterns could shift. Without proxy advisors acting as a common reference point, outcomes may become less predictable. Some executives welcome the change, arguing it could lead to more thoughtful, company-specific engagement. Others worry it could add uncertainty to an already complex governance environment.
Regulators, too, are paying attention. While there is no prohibition on using AI for proxy voting, authorities are increasingly focused on how AI systems are governed, tested, and monitored. Ensuring that automated tools comply with fiduciary duties and do not introduce hidden biases will be a key concern.
For JPMorgan, the move is both a technological experiment and a statement of intent. It reflects a belief that the future of asset management lies in combining scale, data, and advanced analytics to gain greater control over core functions. Whether AI-driven proxy voting becomes the new norm or remains a strategy adopted only by the largest firms remains to be seen.
As the proxy season unfolds, market participants will be watching closely to see how JPMorgan’s votes compare with traditional recommendations and whether this bold shift delivers on its promise. What is clear is that the use of AI in shareholder governance has moved from theory to practice — and the ripple effects could be felt across corporate America.
