AI is here and with it comes anxiety of what the future holds. However, is it really a scary unknown? Looking at the positives I try to see AI not as a threat to the discipline of product management, but an aid for the fundamentals we already practice.
Working as a product manager in various large businesses, I've seen how they are reacting to AI. The size and complexity of these environments creates its own challenges; legacy systems, competing priorities, multiple stakeholders and a long decision making chain. I've found that the fundamentals of good product management don't disappear in this space, if anything, they matter more. Here's what I've learned so far.
In large businesses, this is where things most commonly go wrong. Organisations with the budget and appetite to move quickly can launch AI initiatives that are driven by enthusiasm rather than a clearly defined problem. The result is often a proof of concept that never scales, or a tool that gets adopted without anyone measuring whether it's actually a success. The same discipline a good PM brings to any initiative applies here: define the outcome first, agree how you'll measure it and make sure there's an owner accountable for it.
2. Clear Direction In, Strong Results Out
As with the user stories you create now, and instructions you give to teams, AI is only as good as the information you provide. Be specific about your goal, the context you're working in and how you want the output formatted. The more clarity you provide upfront, the less rework you do later. As has always been the case, as Product Managers we need to give clear direction.
Think of it like writing a brief for a development team. Vague requirements produce vague output, and then you spend time in review cycles fixing what could have been right the first time. With AI, the same principle holds. A well-structured prompt that includes your goal, the relevant context, the audience and the format you need will consistently outperform a vague one-line description. A practical example: if you need to generate user stories, don't just ask AI to 'write some stories for a login feature.' Instead, give it your BDD template: Given, When, Then. Tell it the persona, the goal, and any acceptance criteria you already know. Then how you want that output structured. The skill of prompting well is genuinely learnable and for PMs who are already trained in writing clear, outcome-focused requirements, it's a natural extension of what you already do.
3. Faster Answers, Right Questions
Product Managers need to process large volumes of information to help give business direction and understand success. AI gives you the ability to cut through this far faster than any manual process. However, it still needs a human to interrogate the output, the interpretation and the judgement call remain yours, you can just get there sooner.
Large organisations generate enormous amounts of data, research, documentation, stakeholder feedback and market analysis, and much of it goes underused simply because there aren't enough hours to process it all. AI changes that; it allows access to information far faster, and it surfaces patterns, giving you insight and suggestions. However, it won't know the internal politics, the strategic context, or the nuance that only comes from being inside the organisation. Treat its output as a starting point rather than a conclusion. The risk isn't that AI is wrong, it's that it can be wrong in a way that sounds completely right. When you get an output back, ask yourself: does this contradict anything I already know? Has it missed context that wasn't in my prompt? Is it making assumptions I haven't validated? Apply the same scrutiny you'd give to any internal report or third-party recommendation. Iteration is part of the process, not a sign that something has gone wrong.
New technology can feel overwhelming, and AI is no different, the pace of change is significant and it's easy to feel like you're already behind. Reconnect with your curious side; AI is one of the most useful tools available to you right now, used in the right way, and the best way to get value from it is to just start exploring.
People I've seen get the most from AI are the ones who approached it with a willingness to experiment, trying it on real tasks, seeing where it helps and where it falls short and building their own working patterns from there. In large businesses there's often a tendency to wait for a sanctioned tool or a company-wide rollout before engaging. Start exploring within whatever guardrails your organisation has in place, most will have approved tools or sandboxes you can use. Keep proprietary and customer data out of anything non-sanctioned and build your instincts from there.
AI isn't a replacement for product thinking, it's just another tool at your disposal. The PMs who will get the most from this shift are the ones who bring their existing skills with them: clarity of purpose, critical thinking, and the ability to ask the right questions. Those fundamentals haven't changed. The tools available to apply them have.
If any of this resonates and you're looking at how AI can work harder in your business, reach out to me or the team at Crosstide, we work with large businesses every day on exactly these kinds of challenges.