"Yes, more times than I can count! The tip-off is usually the awkward silence after someone says We should use AI.' Everyone nods, but nobody's picturing the same thing. In my experience working across campaigns and communications, 'AI' in one breath has meant automating admin tasks, in the next it meant drafting copy, and in another it meant analysing audience data for strategic decisions. Three completely different capabilities, three different risk profiles, three different questions about who stays accountable. What I've found works is simply asking: are we trying to create something, automate something, or understand something? That one question cuts through the noise fast. What goes wrong when it all gets lumped together is that projects get scoped around the hype rather than the actual need, and when the tool doesn't deliver the magic everyone imagined, AI gets blamed rather than the lack of clarity upfront. The project manager's role here is crucial: defining what 'AI' actually means in the context of this project, for this team, toward this outcome. Saving Changes...
Frank SpiegelSenior Projectmanager, PMP, PMI-ACP| Commerzbank AGOberursel, Germany
I treat “we should use AI” as a vague symptom or the wish of someone to show to be up-to-date, not a solution. To unpack it, I propose to walk through a structured set of questions in plain language, aimed at getting from buzzword to a concrete, testable problem. You can think in five layers:
When someone says, “we should use AI,” I usually pause and ask, “For what exactly?” Most of the time, they are not really asking for AI - they are asking for a solution to a problem. Maybe decisions are taking too long, reporting is messy, customers are waiting too much, or the team is repeating the same manual work every day. So I try to unpack it by asking:-
What problem are we actually trying to fix?
Where is the delay, cost, or frustration?
What would a better result look like?
Do we already have the right data?
How would we know if this worked?
That usually shifts the conversation from hype to something practical. Sometimes AI helps, and sometimes the better answer is simply a cleaner process, better ownership, or basic automation. Saving Changes...
The most important is to define what AI means. Based on that we will be able to use it on a proper way. Saving Changes...
THANESH ARUMUGAMSenior Project Manager| Hitachi Rail STS MalaysiaRawang, Malaysia
We can frame our answer around saving time and increasing output. Saving Changes...
Robin SullivanDirector Program Management| Advanced Micro DevicesAustin, Tx, United States
We should use AI whenever possible if it adds value. And in most cases, it does. Another thing to consider is your companies governance and availability of tools and access to data. Sometimes, companies don't keep up with the latest or limit access of data. Saving Changes...
I’ve noticed “AI” can mean a lot of different things depending on the person. Sometimes they mean automation, sometimes they mean a chatbot, and sometimes they just want help making sense of data.
What usually stands out to me is when the conversation is focused on the tool instead of the problem. I think it helps to slow down and ask, “What are we actually trying to improve?” When AI gets treated as one big category, it can create confusion and unrealistic expectations. The clearer we are about the task, the easier it is to choose the right solution. Saving Changes...
ANGELA ROCHAPM Specialist| TELEMAR NORTE LESTE SARecife, Pernambuco/Brasil, Brazil
If we can simply improve the process and automate the basics, addressing the problem, we probably won't need to implement AI tools. However, if there's still a need to analyze complex patterns, learn from data, or support decisions at scale, then AI might be justified and its viability tested Saving Changes...
"Compañeros, cuando un stakeholder pide IA, ¿realmente está pidiendo innovación o está manifestando una falta de confianza en la capacidad de entrega (throughput) actual del equipo? Antes de elegir el modelo de lenguaje, yo elijo analizar la causa raíz de esa urgencia." Saving Changes...
This has been a great introduction of understanding to ask the right type of AI questions up front and early. I agree with many responses above that I am observing a lot of folks afraid of AI, which if understood more, in actuality is creating super-powers within each of our own individualities...ACTIVATE! Additionally, I have tools I am still navigating myself, so I am humble that I still have a lot more to learn. The continuous learning opportunities are exciting! I find it easier than having to create my own ground-up information with these AI tools emerging. This is only the beginning and quite an exciting time for everyone.
PMI Team Member
