It seems that a raging debate over whether AI will dramatically change investor communications has been laid to rest. It almost certainly will.
Today, 81% of North American IROs are using AI to summarize peer and market events, including transcripts, research and news, according to the most recent Nasdaq Pulse survey. In addition, Nasdaq has found that 70% of IROs are currently using AI for scripts and Q&A support as part of earnings prep.
Some of the most compelling signs that AI is no longer a pipedream come from key players in the investor community using AI in ways that would have been unimaginable a year or two ago.
Take Gregg Lampf, Vice President of Investor Relations at Ciena in Maryland, who relies on AI for everything from anticipating market reactions to crafting skillful investor emails in sensitive situations.
He believes that using AI is a matter of fighting fire with fire given that his audience isn’t exclusively human anymore. Investors have told him they generate questions – in real time – with the help of AI tools. What’s more, not only are some sell-side reports written by AI, but at least one ETF, Qraft AI-Enhanced U.S. Large Cap Momentum, has tasked machines with portfolio allocations.
Nathalie Megann, President and CEO of the Canadian Investor Relations Institute (CIRI), tells a similar story: “AI is poised to revolutionize investor relations by offering enhanced efficiency, strategic insights and improved communications. AI is here to stay, and it needs to be embraced – with the proper governance and safeguards in hand.”
Understanding AI’s New Role
For those IROs only now getting acquainted with AI, there are a number of simple, safe ways to proceed.
For instance, Megann suggests asking ChatGPT, Claude, Google Gemini, Microsoft Gemini, or any other AI tool to analyze a company’s most recent IR report, examining what specific language choices are most effective.
Another idea is to task AI with summarizing internal meeting notes, highlighting the most salient points.
Many IROs feel safest noodling around on closed systems, such as Perplexity, so “they’re not feeding the systems material, non-public information,” says Alyssa Barry, President of Alliance Advisors IR.
Barry emphasizes that AI has freed up time so she can undertake projects on the back burner. As an example of time savings, she estimates that using AI slashes the hours she spends drafting data-focused case studies by a whopping 60%.
Canadian IROs report similar experiences. For Ivan Zarate, Principal, Investor Relations, at Air Canada, “using Copilot is like having an analyst by your side 24/7.” He notes that AI has shaved off at least five hours from writing and preparing an MD&A.
What’s more, Zarate no longer reads every sell-side analyst report or the transcript of competitors’ calls in their entirety because AI can summarize the content so fast and so well.
Greg Secord, Vice President and Global Head of IR at Open Text Corporation, uses AI to pull together up-to-the-minute stats on macroeconomic issues. “In a world where every 90 days the macro environment changes and people’s questions change, too, we want to have the answers at our fingertips,” he says.
Down the road, Secord envisions AI acting as “’a test-case’ investor and commenting on PowerPoints, releases and earnings scripts.” In this scenario, AI might identify questions that haven’t been properly addressed or points that aren’t sufficiently clear. “We’re not 100% there yet, but a scenario like this would be very exciting,” he says.
Using AI to Revamp Earnings Calls
Lampf considers AI’s potential to anticipate investor questions a game changer. Recently, he’s delved deeper and begun asking AI which questions to expect from different segments of his investor base. For instance, he’s asking AI to identify questions a bull or bear might ask – or how an announcement might land differently with a growth or value investor.
During this spring’s earnings call, just after Ciena was added to the S&P 500, Lampf wrestled with how best to introduce his company to a broad range of new investors. He took a stab at drafting language, then asked AI for feedback.
“I don’t dump information into AI and let it generate a script,” he explains. “I’m more comfortable the other way around. We write what we want to say and then ask AI to review.”
He continues: “We’re very much human-led users of AI. After all, we know our business better than AI does and we want to have the right voice.”
Others – like Jeff Robinson, Sales Director, Canada, for Lumi Global – are using AI for evaluating the success of investor events. Instead of scouring Reddit and the investor boards himself, Robinson lets AI generate a summary of the prevailing sentiment. He then devises a post-meeting narrative that’s responsive to the mood of the moment.
Attracting AI Attention
Q&As on IR websites have roared back into relevancy because of what Chris Makuch, Notified’s Senior Director, Sales Canada, is calling “the answer-engine economy.” He explains that influencing large language models (LLMs) is now critical. To succeed, IROs must learn to anticipate the key questions investors and other stakeholders will ask about their company – and then feature those questions alongside comprehensive answers on their IR sites.
Just as most IROs didn’t tackle SEO on their own, tools are being created to help with AEO, or answer-engine optimization. On April 29, for instance, Notified unveiled its AI Press Release Optimizer to help communications professionals improve their visibility with LLMs. Similar tools are available through vendors such as PR Newswire (Amplify) and through Access Newswire (AI Toolkit).
Makuch notes that LLMs typically begin by scouring an IR site for answers; if the LLM comes up short, it looks elsewhere – with unpredictable results. “I’ve plugged questions into Google and the source I’ve gotten is Wikipedia,” he says. “I guarantee you there isn’t an IRO out there who wants the source document for a question about their company to be Wikipedia.”
While many IROs are honing in on the potential time- and cost-savings provided by AI, “ensuring that the source that AI and the LLMs are using is your website, your corporate story, and your press release” is of the utmost importance, says Makuch. In other words, he explains, making your investment case in terms that bots will find and deliver to a wider audience is the crux of IR messaging in the AI era.
Agentic AI: Trusting the Bots
Nearly everyone has interacted with an AI chatbot on an airline or hotel site, yet practically no public companies offer the same technology to investors. This observation led Nick Kuzyk, CEO and Founder of Calgary-based Answir, to build an AI assistant exclusively designed to respond to investor questions. Answir stands out for being built exclusively to answer IR questions, and yet other well-known IR platforms can do some of the same.
Kuzyk estimates that the average public company could use AI to handle 80% of investor queries. Answir supplies investors with EDGAR and SEDAR+ information, stock data from the exchanges, voluntary uploads and company-sanctioned web sources. When addressing IROs’ qualms, he contends that a company-sanctioned AI assistant capable of differentiating between authorized sources of information and Reddit-style rumours is vastly preferable to other AI options.
Megann recognizes the appeal of chatbots supplying IR information so long as the information has been vetted and complies with all regulatory requirements.
Additionally, she cautions IROs that “chatbots are limited to regurgitating what’s been written and put out into the universe.” She continues: “Today’s chatbots just aren’t capable of talking about the rationale behind a strategy and meaningfully discussing risk factors.”
Over the coming months, Lampf plans to investigate using AI agents on Ciena’s IR site. “First, I want to know how these agents operate and what’s in their DNA,” he says. “But I do think there are going to be agent-to-agent environments. I think that’s where AI is going next.”
What Lies Ahead
Those interviewed for this article no longer fear that AI will take their jobs; instead, they’re concerned that not knowing AI is a sure path to obsolescence.
Another prominent theme was the folly of leaving AI to its own devices. Human oversight and personal relationships still matter keenly.
“No matter how you use AI, there always has to be a human element to it,” emphasizes Megann. “Everything has to be scrutinized, and someone held accountable for whatever information you send out to the universe.”
While intriguing use cases abound, IROs are currently intent on harnessing the potential of AI to streamline writing, data-gathering and other research-oriented tasks. Doing that, they argue, frees more time for strategy and one-on-one engagement.
“What sets us apart [as IROs] is the relationships we’re building with the Street, with each other and with people inside our organizations,” concludes OpenText’s Secord. “If we can make tools that make life more efficient and that make us smarter about the markets and about our companies, then there’s a real advantage to being AI literate and working in an AI environment.”