For me, the overarching limit is "Don't implicitly trust AI." The presents itself in at least three areas: 1) don't let it make decisions for you, 2) don't give it any sensitive or confidential personal or business information, and 3) challenge the response. This can look like:
- Challenge the problem framing. AI answers the question it thinks you asked, not what you're trying to ask. - Expose hidden assumptions. AI doesn't tell you when it's filling gaps - ask it to identify the assumptions it's making. - Ask for supporting data, evidence for claims, and sources of reasoning - Test edge cases and boundary conditions - Look for overgeneralization - the response may not apply in all circumstances - Cross-check with alternative perspectives - AI often gives the most commonly accepted view; it's not always the correct view - Ask AI to separate "what is" from "what should be". It can blend the two without telling you. - Check for internal consistency - When obtaining estimates, ask for confidence levels and uncertainty ranges
Practical limits should be based on reasonable levels of trust and understanding its capabilities - a combination of what it can do, what it can't do, and what you shouldn't let it do for you regardless of whether it CAN do it. For example, you CAN have it generate your user stories or use cases. It may identify things you wouldn't, but it you don't have an internal expert review them you also risk producing irrelevant, incomplete, or over-developed deliverables. AI can inform, but a person needs to decide. Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
No limit. We are using AI from more than 40 years ago sometimes without notice that because it is embedded into our tools. The break was when in 2017 a new subset of AI emerges: generative AI. In this case to reinforce all related to Responsible AI is a must. Saving Changes...
Program Manager| HARPER SRLSanto Domingo / Distrito Nacional, Dominican Republic
I don’t think there’s a fixed “limit,” but there are clear boundaries. AI is useful for supporting work like drafting, summarizing, organizing information, exploring options, and speeding up repetitive tasks. It can help you move faster and see things you might miss, especially in early analysis or documentation.
Where it shouldn’t be relied on is in final decisions that impact scope, cost, people, or ethics. Anything that requires full context, judgment, or accountability still sits with the PM. At the end, the limit is responsibility; if you’re accountable for the outcome, you can’t delegate that to AI. Saving Changes...
AI is powerful in projects, but it has a sensible boundary. For a project manager, the limit is simple: use AI for speed, insight and automation, but never outsource judgment, relationships or accountability. AI can read data, suggest risks, draft documents and analyse patterns. It cannot understand team emotions, business politics, stakeholder trust or long term consequences the way a human does.
So the limit is reached when the decision requires empathy, context from lived experience or responsibility for people. AI supports your mind, but it should not replace your role as the guide of the project. Saving Changes...
AI doesn’t really have a hard “limit” in terms of where it can be applied in project management, such as planning, risk analysis, reporting, communication, and more. The real boundary isn’t capability, it’s ethical and responsible use. As a project manager, you need to ensure AI is used in ways that protect data privacy, avoid bias, maintain transparency, and keep humans accountable for decisions. AI should support judgment, not replace it.
So instead of “the sky is the limit”, a more grounded view is: AI can be used widely, but always within clear ethical, legal, and organizational boundaries. Saving Changes...
Product Operations Program ManagerBarcelona, Cataluña, Spain
As with many things in life, finding balance is key. It is important to analyze which tasks can be performed more efficiently with (generative) AI and which ones still benefit from human intelligence and common sense. Saving Changes...
Senior IS Project Manager| Baycare Health SystemsClearwater, Fl, United States
I agree with Aaron's post.: "For me, the overarching limit is "Don't implicitly trust AI."
As a Project Manager, you need to be sure to not overstep your organization's governance regarding the use of AI tools. Some organizations have specific AI solutions that you can use, and everything else if off limits. Saving Changes...
"We cling to our own point of view, as though everything depended on it. Yet our opinions have no permanence; like autumn and winter, they gradually pass away."
Community Champion
