There’s a lot of noise around AI replacing jobs. In asset management, that concern tends to show up in quieter ways — less about eliminating roles, more about uncertainty. What will change? Where will value shift? What does this mean for how teams are structured?
The more useful question isn’t about replacement. It’s about what AI reveals.
Because when firms begin implementing AI tools seriously, something consistent happens: processes that felt manageable before suddenly feel inefficient. Not because anything changed in how the firm operates, but because AI forces a level of clarity that surfaces problems that were already there.
What AI Actually Exposes
Here’s what we see at smaller asset managers when AI implementation gets underway:
Investor reporting
The process takes four to seven business days, involves three or four people across finance and IR, and requires manual reconciliation between the fund admin data, the internal model, and the narrative template. Each cycle, someone finds a number that doesn’t match. Nobody is sure which version is current until the partner reviews it.
AI can’t accelerate this process until the data problem is solved. And solving the data problem requires acknowledging, often for the first time, that the reporting workflow is more fragmented than anyone realized.
DDQ and RFP production
The firm has answered the same 40 questions in slightly different ways across 60 different DDQs over three years. The answers live in email threads, in old Word documents, in the head of the IR associate who assembled most of them. When that person leaves, the institutional knowledge goes with them.
AI can dramatically reduce the time it takes to draft DDQ responses, but only once there’s a clean, trusted library of answers that someone owns and keeps current. Building that library forces the firm to confront how much undocumented knowledge exists in a handful of people.
Portfolio monitoring and reporting
Data comes in from portfolio companies in different formats, on different schedules, with different levels of completeness. Someone on the ops team spends a significant portion of their time reformatting, chasing, and reconciling before anyone can do actual analysis.
AI agents can automate significant portions of this: document extraction, flagging of anomalies, and initial formatting. But the workflow upstream of the automation has to be clarified first: what format is expected, what fields are required, what happens when data is missing.
In each of these cases, AI doesn’t create the problem. It makes the problem visible in a way that’s harder to work around.
The Inefficiency Multiplier
This is the pattern we call the inefficiency multiplier: AI amplifies whatever is already present in the workflow.
- If a process is inconsistent, AI produces inconsistent outputs.
- If ownership is unclear, AI adoption stays optional.
- If the data isn’t trusted, the outputs won’t be either.
- If the workflow depends on a few key people, AI makes that single-point-of-failure risk visible (often at the worst time.)
This is why some firms feel like AI is underdelivering. It’s not that the technology isn’t capable. It’s that the technology is revealing structural issues that were already there, and now they’re harder to ignore.
The Talent Gap That’s Emerging
The firms that are getting ahead of this aren’t replacing people. They’re adding a new kind of person, and that person is genuinely hard to find.
The profile that’s in demand right now at asset managers who are serious about AI is not a data scientist and not a traditional operations manager. It’s something in between: someone who understands the business workflow deeply enough to redesign it, understands data well enough to know what’s usable and what isn’t, and understands technology well enough to work closely with engineers and system vendors — while staying accountable to the business outcome.
These roles don’t fit neatly into traditional job descriptions. They’re not showing up on org charts yet at most firms. But they’re the roles that make the difference between AI that stays in pilot mode and AI that changes how a team actually operates.
The challenge is that most candidates are strong on one side — they’re either deep operators who are curious about technology, or technically skilled people who lack the financial services context. The genuinely hybrid profiles are rare, and they tend to move quickly.
We’re actively placing these profiles at asset managers right now — across data foundation roles, workflow transformation leads, and AI enablement functions. The work of identifying them requires looking well beyond the job title on the resume.
What This Means for COOs and CFOs
For COOs:
The question to ask is: which of your current workflows would break down fastest if your most knowledgeable person left tomorrow? That’s where AI exposure tends to be sharpest, and where the documentation and ownership work needs to happen first.
For CFOs:
The question is: which of your recurring processes still require a senior person to review and validate every cycle, not because the judgment is complex, but because the underlying data can’t be trusted without human verification? That’s where AI readiness work pays off fastest, because it solves both the efficiency problem and the data governance problem at the same time.
The Real Question
The question shouldn’t be: ‘Will AI replace our team?’
The more useful question is: ‘What will AI reveal about how our team operates today, and are we set up to act on what it shows us?’
Because that’s where the real transformation begins. Not in the technology. In the operational clarity that the technology forces you to achieve.