Recap -- Beyond the Prompt: Mastering the Intersection of AI and CI
- Mark Gigliotti
- Apr 23
- 3 min read
We recently hosted a webinar on AI and competitive intelligence. Before I share some of the big takeaways, I want to note for the record that it is April 2026, because this conversation will mean something different in six months. That's just the world we're operating in right now!
Joining me was Glen Remy, Director of Marketing and Competitive Intelligence at Nice, who has been in CI for a long time and brought a grounded, practitioner perspective to the whole conversation. Here's what stuck with me.
AI changed the balance of work, not the nature of it
The way Glen framed it was simple and I think accurate: AI helps find the needle a little faster. The haystack is still there. What has shifted is where analyst time goes. Gathering information used to be the bigger slice of the pie. Now it's a smaller slice, and verification has gotten larger.
One thing worth saying plainly: AI in CI is still secondary research. It is super-powered secondary research, but it is public data that a model can connect and summarize faster than any team could manually. It is not going to give you what is inside someone's head, what a customer said off the record, or what a competitor's implementation really takes versus what their sales team claims. That distinction matters.
The sales rep problem is real
Glen mentioned something that I think a lot of CI teams are quietly dealing with. Sales reps are building their own competitive materials using ChatGPT or whatever LLM they prefer, and increasingly they are not running those materials past CI before using them. A year ago, people at least sent the output over and asked someone to validate it. That step is starting to disappear.
This is not a new problem. Sales people have always found ways to develop their own materials. But now they have a tool that produces something that looks authoritative and complete, and the confidence gap between what the AI outputs and what has actually been verified is not visible in the slide.
At Nice, Glen's team responded by building guidelines around AI use rather than trying to stop it — if you are going to use it, here is how to do it in a way that does not get you into trouble. That feels like the right approach.
The knowledge repository is the actual competitive advantage
This came up a few different ways throughout the session, but the core point is that LLMs only have access to what is public. The next level down — primary research, win loss interviews, validated field intelligence, the things your CI team has actually gone out and gathered — that is not in the model. That is yours.
The teams that are going to get the most out of AI-assisted CI are the ones who build a validated internal repository that an LLM can draw on alongside public sources. Without that, you are just getting a well-organized summary of what is already out there, which your competitors can get just as easily.
A good battle card has what Glen called "the dirt" — hidden costs, implementation timelines that don't match the pitch, features that exist on paper but are clunky in practice. None of that is on a vendor's website. Getting it requires a human being asking the right questions of the right people.
The CI role is not going away >> it's shifting toward consulting
If battle card automation gets easier, and it will, the question is what CI professionals do with the time that frees up. The answer Glen and I landed on is consulting — getting on calls with sales teams around specific deals, bringing the validated intelligence that is not in the public AI output, and being the person in the room who can say, yes, this is true, we have seen it, or no, that claim does not hold up.
I think it is time for some swagger in the CI industry on this point. There is a layer of work that requires human judgment, validated sources, and real accountability for the intelligence you are putting into the world. That layer is not going away. If anything, as AI-generated content becomes harder to distinguish from verified research, the value of someone who can say "we know this is true and here is how we know" goes up.
The companies that recognize that will be better positioned. The ones that hand everything to an LLM and skip the verification step will eventually get burned by it.
If you work in CI and want to talk through how your team is navigating any of this, reach out. We work with organizations across tech, CPG, industrials, and financial services on primary research and competitive intelligence programs.
