Is Investor Relations Being Left Behind in the AI Push?

(this article was written with the help of Fiona Sherwood, founder of MarketMy.ai)

Just about every alternative asset manager is using AI. The question is whether they're using it in all the right places.

 AIMA reports that 95% of alternative fund managers are now using AI in some capacity.  What that number doesn't tell you is what for. For most GPs and hedge fund managers, the AI budget is flowing in one direction: investment research, deal sourcing, and portfolio management. The analysts, portfolio managers, and deal teams are getting resourced. The rest of the firm is still on the sidelines.

The Investor relations function, though, is one area ripe for efficiency gains.  At its heart, investment management is a relationship business, and IR teams spending less time filling out RFPs or putting together monthly tear sheets from using AI-enabled tools will be a win for everyone involved.

But the firms that will benefit from AI aren't just the ones that buy the right tools; they're the ones that deploy them with compliance and governance in mind. Policies that hold up under LP operational due diligence, take a knowledgeable approach to generic versus purpose-built solutions, and demonstrate a clear adoption strategy will separate the firms that gain a genuine operational edge from those that add another underused platform to the stack.

Where AI Moves the Needle

There are numerous use cases where sales, IR, marketing, and distribution teams can optimize their workflows using AI tools beyond the generalist LLMs most people are becoming familiar with, e.g., ChatGPT, Claude, Gemini.  Many of these purpose-built tools have been designed to facilitate capital raising while improving the speed and accuracy of investor reporting.  The tools exist, and the firms that put them in the hands of their IR and distribution teams will have a measurable advantage over those that don't.

DDQ and RFP Automation. 

Completing institutional DDQs manually drains senior IR time across every type of manager and strategy. Purpose-built tools can pull from existing document libraries to draft responses, flag gaps, and maintain version control. The time savings are real, and the compliance benefits - consistent, auditable, version-controlled responses - are equally significant.

Capital raising Intelligence.

 Predictive analytics can identify which LPs are most likely to make a commitment, which prospects are actively allocating to your strategy, and where capital is moving across channels. That's not a replacement for relationship judgment; it tells you where to maximize the relationship time you actually have.

Tools like AI-enabled CRMs provide the distribution infrastructure that sales teams in other verticals have used for years. The wealth channel in particular demands it; the sales motion is different, the volume is higher, and hand-holding expectations are greater.

LP Onboarding.

Getting investors over the finish line is a friction point that too many firms accept as a given. Subscription documents, KYC workflows, and AML checks are time-consuming by nature, but losing committed capital because the closing process is slow or cumbersome is an entirely avoidable problem. AI-enabled onboarding platforms can automate document preparation, pre-populate investor data across multiple funds, flag compliance issues before they cause delays, and reduce the back-and-forth that drags out tight timelines. For LPs investing across multiple managers, the firms that make this process frictionless stand out. The technology exists. The barrier now is simply adoption.

LP Reporting and Investor Portals. 

This is where the pressure is most visible right now. Distributions are down, IRRs are under pressure, and LPs are demanding faster, more granular transparency. AI-assisted portfolio monitoring tools aggregate and synthesize portco data - financials, KPIs, management commentary - compressing the time from data collection to investor-ready output. AI tools can generate first-draft commentary directly from the underlying data, which IR refines rather than writes from scratch.  For hedge funds, the reporting burden is equally acute but shaped differently. Investors expect timely, accurate performance attribution, risk exposure summaries, and factor analysis, often on a monthly or even weekly cadence. AI can automate much of the data aggregation and formatting that sits behind those deliverables, reducing the operational lift on IR teams while improving consistency across investor communications.

Getting the data right is half the battle. How it reaches the investor is the other half.  Most LP portals are still document dumps with a login. AI-driven portals with personalized dashboards, natural language queries, and proactive alerts change that dynamic. But the more urgent point is this: LPs aren't waiting for their managers to catch up. Institutional investors are deploying their own AI tools to ingest and analyze what their managers send them, most of which still arrives as a PDF. For GPs, that's a missed opportunity. Delivering clean, structured, timely data isn't just a service improvement; it's a chance to shape the narrative, deepen the relationship, and demonstrate operational credibility at exactly the moment it matters.

Why Most Firms Are Still Stuck

The use cases are well established. The tools exist, they work, and the ROI case for deploying them in IR is not hard to make. The gap between what's possible and what's actually happening inside most alternative asset managers isn't a technology problem. It's an implementation problem, and it has three distinct components.

The first is governance.

Very few alternative managers have an AI policy that would survive scrutiny in an operational due diligence review. That's a problem that compounds quickly. LPs are increasingly asking about AI governance in DDQs - how models are used, what data they touch, and who has oversight. A firm that can answer those questions with something more than a vague commitment to "responsible use" will stand out. A firm that can't will find those questions increasingly difficult to deflect.

The second is shadow IT.

Where policies are unclear or overly restrictive, employees don't stop using AI tools, they find ways to use them on personal devices instead. The compliance implications are significant. Sensitive LP data, fund information, proprietary deal intelligence: when team members are feeding this into consumer-grade LLMs outside any governance framework, the firm has a problem it may not even know it has. The answer isn't to lock everything down - it's to create clear, workable policies that give people sanctioned tools and the training to use them well.

The third is tool selection. 

Most firms are still navigating the tension between generic platforms and purpose-built solutions, and the decision matters more than many appreciate. Tools like Microsoft Copilot can deliver real productivity gains across a team with shared agents, streamlined workflows, and systematized processes. But Copilot and its generalist equivalents aren't built for the depth of retrieval that IR work demands. When an IR professional asks Copilot to search across SharePoint for RFP responses, it will surface something. It won't necessarily surface the right thing, because it's reading at breadth, not depth.

Purpose-built platforms use retrieval-augmented generation (RAG), which means they're drawing on deeply indexed, domain-specific knowledge bases rather than skimming a broad environment. For DDQ responses, performance commentary, or compliance-sensitive communications, that distinction isn't a technicality. It's the difference between a draft you can work with and one you have to rewrite from scratch.

The right answer for most firms isn't one or the other. It's understanding which jobs each type of tool is actually suited to, and being honest about the gaps.

The Adoption Problem Is a People Problem

Technology selection gets most of the attention in these conversations. Team readiness gets almost none, and that's where most implementations quietly fail. Buying a purpose-built IR platform or rolling out Copilot or Claude across the firm doesn't automatically change how people work. It changes what's available to them. Whether they use it well, consistently, confidently, and in ways that actually reflect best practice depends almost entirely on how the rollout is handled. That means training, not just onboarding. It means addressing the anxiety that sits underneath most AI adoption conversations, even when it isn't voiced directly. People worry about getting it wrong. They worry about what it means for their role. They worry about trusting an output they can't fully interrogate.

Those concerns are legitimate. The firms that acknowledge them and build adoption programs that address them directly will get materially better results than those that send a system access link and call it a day.

There's also a capability gradient problem that most firms underestimate. Within any IR team, you'll have people who have been experimenting with AI tools for two years and people who are still largely avoiding them. If you deploy a new platform without accounting for that spread, you get uneven adoption at best and active resistance at worst. The infrastructure matters, and so does the approach to bringing people along with it.

What Good Looks Like

The firms getting genuine value from AI in IR tend to share a few common characteristics.

They've done the governance work first. Not as a compliance exercise, but as a foundation for adoption. They have identified clear policies on what can be used, for what purpose, with what data, and with what human oversight. This makes it possible to deploy tools confidently and to answer LP questions about AI usage without scrambling.

They've thought carefully about where generic tools end, and purpose-built tools begin. They're not wedded to one approach or the other; they've mapped their actual workflow requirements and made decisions accordingly. And when they deploy, they deploy with proper change management, like training, communication, and iterative rollout, rather than assuming the technology will sell itself internally.

Perhaps most importantly, they've recognized that AI in IR isn't a project with an end date. The tools are evolving, LP expectations of the information needs are evolving, and the operational standard for what good looks like will continue to shift. The firms treating this as a capability to build over time, rather than a system to install once, are the ones that will maintain the advantage.

The Relationship Isn't Going Anywhere

There's a version of this conversation that positions AI as something that could eventually replace the IR function. That's not where this is going, and most experienced practitioners know it.

Alternatives is a relationship business in a way that other asset classes simply aren't. The commitments are larger, the due diligence is deeper, the lock-up periods are longer, and the stakes of a deteriorating LP relationship are proportionally higher. The trust that sits at the center of that dynamic, between GP and LP, and between IR professional and investment committee, isn't something a model can replicate.

What AI can do is protect that relationship from being undermined by operational friction. Slow reporting. Inconsistent DDQ responses. Portals that are difficult to navigate. Onboarding processes that drag. These aren't trivial irritants; they're the kind of persistent experience failures that quietly erode confidence over time, particularly when an LP is also dealing with compressed returns and longer hold periods.

The firms that use AI to eliminate that friction, and do it in a way that's governed, transparent, and embedded in a genuine adoption strategy, are creating more space for relationship building, which is the true driver for AUM growth.