A perspective from Annie Eissler, CMO, Altvia
I recently moderated a panel at PEI’s IR Network meeting in Chicago and spent the rest of the afternoon in roundtables and side conversations with IR professionals across the industry. While the official topic of my panel was AI and technology in IR workflows, AI ended up being the dominant theme in every panel and conversation throughout the day.
What struck me most were the incredibly thoughtful conversations, insights, and concerns that were discussed throughout the day. The real conversation wasn’t whether AI matters, it was how teams are actually using it without letting the tool do the thinking for them.
Here are a few of my key takeaways from the day:
No definitive playbook yet, but significant experimentation
What came across loud and clear was that teams are experimenting, comparing notes, running pilots, and making judgment calls in real time. But the pace of change in AI right now is often outrunning internal approval cycles, compliance reviews, and honestly, most people’s capacity to keep up.
Several attendees who felt optimistic and comfortable about their team’s AI usage had one thing in common: clear guidelines and guardrails in place. Some examples included:
- Building tools that can be deployed across the firm, as opposed to individual prompting.
- Specific processes that include where tech is used and where human review/touchpoints are along the way.
- Establishing a rule: AI used for operational leverage, but all communications need human eyes.
LP attitudes on AI are mixed and that should shape how you use it
One of the more clarifying moments of the day came when LP attitudes on AI came up directly. The feedback was genuinely split. Some LPs say they don’t mind knowing AI is being used in the communications they receive, while others said they identify it in the first few words and move on.
But here’s what was consistent across the IR side of the room: nobody wants their LP thinking “AI wrote this.” Not because AI can’t produce good output, but because the moment an LP feels like they’re on the receiving end of a template, the relationship takes a hit.
However, I urge teams not to fall into the trap of trying to hide that AI was used, and to remember that the goal is to make sure what you’re sending is relevant, truthful, and specific enough to signal that you actually know who you’re talking to. Attendees all agreed that communications should be brief enough to respect LP’s time, and human enough to warrant a response.
The human element is non-negotiable
What came through clearly on the IR side is that no matter where LPs land on AI-generated outreach, IR professionals don’t want to be in a position where an LP feels like the relationship is being managed by a bot. The reputational risk is real, and it’s shaping how carefully teams are thinking about what AI touches and what it doesn’t.
AI can handle the prep, the research, the first draft, the operational lift. It cannot build a relationship, and it can’t do business.
There was a story that surfaced in the room about an IR associate who used AI to fix a problem in some analysis. When asked how it was fixed, the answer was “I don’t know, Claude did it.” That’s a risky sentence, and reiterated the need for firm-wide guidelines about where human judgment stays in the process.
Compliance is top of mind
SEC regulations and compliance considerations are sitting in the back of everyone’s mind when it comes to AI in LP communications. Nobody had clean answers. But the awareness is there, and it’s shaping how cautious teams are about what they automate and what they don’t. That guardrail question isn’t resolved yet for most firms, but continues to be a top priority.
Build a tool, not just a habit
One perspective I heard and gave me pause was that there’s a difference between using AI as an assistant and building AI into your infrastructure. Meaning, an assistant answers the specific question posed, but if you build a tool it builds and refines over time. IR professionals should be thinking about how to make their AI capabilities repeatable and institutional, not dependent on an individual’s prompting prowess.
The operational gap is real and unglamorous
This theme came up here, it also came up in our recent webinar with Private Markets Group as well, and it’s boring and unglamorous – data sources, data integrity, and data completeness remains an underlying issue.
Reporting, deal stages, getting notes from a multi-person LP call into the CRM with the right context attached all came up as workflows people want to solve. Excel is still everywhere for portfolio monitoring, and most teams are on CRMs that weren’t built for private markets. The tools people want exist in pieces.
A cohesive AI-native solution for an IR team, end to end, doesn’t really exist yet as a product in this market, and teams continue to look for tools that streamline data and operations.
Relationship building at the center of the conversation
Across every conversation, the through-line cut to the heart of the IR function – how do you build, maintain, and strengthen relationships. Common knowledge is not to show up in front of an LP only when you need them, but to maintain consistent, low-stakes engagement like sharing something relevant or following up after an event.
Of note: short-form content is resonating well. Firms are finding that consistency and thoughtfulness matter more than production value.
The focus has shifted to the institutional relationship, not just at the individual contact level. Where AI entered this chat was in a question posed: “How can AI help me operationalize these consistent, but personalised, engagement points?”
At the end of the day: What does real AI usefulness look like?
One of the reasons I enjoy attending the PEI networking events is that the size lends itself to real conversations about what’s happening across the board in the IR function. If I had to summarize the day in one sentence, it’s that Investor Relations remains a relationship business. Any successful tool or technology that is implemented will make LP engagement feel more personal, not less, but will alleviate operational busywork on the backend.
Nobody has the full playbook yet, but the questions the industry is asking are spot on.
At Altvia, those are the questions driving our product direction.
We build LP and deal sourcing engagement software for private capital firms. If you’re thinking through what AI-enabled IR looks like in practice, we’d like to be part of that conversation.
Schedule time to talk to our team.