Michael Oak
The conversation related to artificial intelligence in institutional investing tends to default to one of two extremes: Either the technology will fundamentally transform the industry overnight or it is an overhyped tool that will fade when the next shiny thing comes along. Neither framing is particularly useful. What is actually happening is far more nuanced, and the organizations that understand that distinction will come out ahead.
AI is not replacing investment judgment. It is exposing weaknesses in data infrastructure, governance practices, compensation design and talent strategy. The more important question is whether an organization is operationally prepared to manage and govern these tools responsibly. And that starts with people.
Most AI Pilots Fail. Here’s Why That Matters for Your Office.
The early data on AI adoption reinforces what many CIOs are already observing: Roughly 95% of enterprise AI pilots generate no meaningful return. In most cases, the tools themselves are not the problem. The failure points tend to occur around the tools. The points of failure include data that are fragmented, unclean or otherwise poorly structured; unclear ownership of systems and workflows; leadership teams that approve a technology initiative without establishing the operating discipline required to support it; and uncoordinated efforts across and within teams on how tools can be most effectively leveraged.
For investment offices, this is not an abstract, scholarly finding. Before expanding into AI-assisted due diligence, performance analysis or board reporting, organizations should start with a more basic question: Is the underlying data environment truly ready for it? Can an investment office confidently say its data are structured, reliable and governed well enough that an AI system can use it without introducing new risks? If most investment leaders answered that question candidly, the answer today would likely be “not yet.”
Where AI Earns Its Place and Where It Doesn’t
The good news is that there are existing practical applications of AI that can deliver real operational benefit. Review of legal documents, manager diligence materials, performance reporting and board preparation all involve large amounts of time spent assembling information. AI tools can help compress portions of that workload and allow investment professionals to focus more of their time on data interpretation, judgment and decisionmaking.
The interpretive layer of investment work remains fundamentally human.
Understanding the data story within the context of an institution’s mandate, stakeholder relationships, risk tolerance and governance requires judgment that cannot be delegated to a large language model. AI can–with good data and instructions–uncover patterns or summarize information. It cannot make the decisions that follow.
For that reason, the CIO role is unlikely to diminish as use of AI tools expands. If anything, the responsibility attached to that role becomes more important.
Where the change will be more visible is at the analyst level. As AI tools get better at aggregation, summarization and scenario modeling, the more mechanical aspects of early-career investment work could likely shrink. Investment teams that begin thinking now about how junior professionals should develop in the new AI environment will be better prepared than those that wait.
Your Compensation Structure Is Either Helping or Hurting
If your incentive structure primary rewards past investment performance against policy benchmarks, it is quietly telling your staff that the hard work required to make AI adoption succeed is not worth their time. This includes everything from improving data quality, piloting workflows, building governance muscle and training teams. If those activities are not recognized in the reward structure, they naturally receive less attention.
Disciplined experimentation and institutional learning are difficult to capture within a backward-looking performance formula. Yet those activities can contribute meaningfully to an organization’s long-term success. A professional who spends extensive time rigorously evaluating an AI tool and documenting why it will not work has still generated valuable knowledge. Most compensation systems have limited ability to acknowledge that contribution.
The talent challenge is compounding. As demand for AI-skilled professionals rises across industries, investment offices must ask whether their compensation programs are competitive enough to attract top talent in one of the hottest labor markets in recent memory–for people with the in-demand skills. Organizations must balance that external pressure with fair pay for existing staff—a tension that will only intensify.
The right mindset here is “get rich slow.” Pay programs that allow room for thoughtful discretion—while offering meaningful opportunities for long-term wealth creation tied to truly exceptional long-term investment performance, will ultimately prove the most effective. Meaningful innovation rarely emerges from environments where deviation from established practices carries disproportionate career risk. Innovative ideas don’t come from organizations that discourage thoughtful risk-taking.
Compensation structures that allow for measured discretion, maintain rigorous performance standards and incorporate longer-term evaluation periods can help create space for responsible experimentation.
No one can predict which investment offices will lead the next decade. But organizations that discourage innovation will almost certainly not be among them.
Talent Strategy Determines Success
In many ways, the talent question is the most important one. Organizations that ultimately derive real value from AI will be the ones that develop internal capability to understand how those tools perform in practice including and especially their real limitations.
As AI changes how work gets done in the investment office, it also demands change to the skills which organizations need to hire, develop and reward. Investment expertise will remain essential. But many roles will benefit from professionals who also possess a deep understanding of data and analytics—and who are comfortable evaluating the output of models critically. Not everyone needs to be a quant or a software engineer. But professionals reviewing AI-assisted work should understand enough to question it, rather than accept it unconditionally.
Organizations that fail to update hiring criteria, onboarding processes and development pathways risk building teams around outdated assumptions that no longer reflect the operating environment.
The Right Question for the Board
The CIO role is not disappearing. Judgment, accountability and the ability to make decisions under uncertainty remain deeply human responsibilities. What will change is the environment surrounding that judgment.
For boards and investment committees, the most important question today is not whether the organization should experiment with AI; it is whether or not the institution is prepared to govern what comes with it.
AI will change how work gets done in the investment office. But the underlying drivers of success have not changed. The organizations that get this right will look familiar: the right people, the right tools and incentives aligned with long-term outcomes. Technology will matter—but the real advantage will still come from the people.
Michael Oak is a managing director of Michael Oak Advisors.
This feature is to provide general information only, does not constitute legal or tax advice, and cannot be used or substituted for legal or tax advice. Any opinions of the author do not necessarily reflect the stance of Institutional Shareholder Services or its affiliates.
Tags: Artificial Intelligence, data analytics, Governance, Investment Management, investment offices, investment staff