To unpack a generic request to "use AI," project managers must first identify whether the underlying intent is task automation, predictive decision support, or generative content creation, as each carries distinct technical and risk profiles. Misalignment typically becomes obvious when different stakeholders use the same label but measure success by conflicting metrics, such as immediate cost reduction versus continuous innovation. Re-anchoring the discussion on the specific business problem and defining explicit human accountability ensures AI serves as a targeted solution rather than a vague mandate. Saving Changes...
Cornel Almondproject manager| officialAppsSan Diego, USA
In project management discussions, “we should use AI” can mean very different things. One person may want help drafting status reports, another may want to predict schedule delays, and someone else may simply want automated reminders. The clues are the expected output, the data required, and how much judgment the task involves. Drafting content, forecasting outcomes, and following fixed rules each call for different approaches—and some tasks only need basic automation. When everything gets lumped together, expectations become unclear, the wrong tools get selected, and teams may overlook accuracy, privacy, or human review. My starting question would be: “What specific problem are we trying to solve, and what would success look like?” That helps turn a broad suggestion into something the team can evaluate. Saving Changes...
Anderson SimonsenGestor de Projetos| Construmaq Construções e MáquinasSÃO LUÍS, Brazil
Quando alguém diz “devemos usar IA”, a organização não deve começar primeiramente escolhendo uma ferramenta, mas compreendendo o valor que pretende gerar. A análise deve considerar se há necessitadamente personalização de produtos, então o sonhado serviços ou experiências, de identificar padrões, tendências ou anomalias em grandes volumes de dados, extrema preleção de falhas, riscos ou comportamentos futuros, de apoiar decisões estratégicas com frequentes inteligência baseada em dados, de automatizar atividades repetitivas e processos operacionais, de melhorar a interação entre pessoas e sistemas por meio da linguagem natural, ou de permitir que sistemas atuem de forma autônoma dentro de limites definidos pela governança organizacional. A partir dessa compreensão, deve-se avaliar a disponibilidade consequente dos dados, a viabilidade técnica, os riscos, os aspectos éticos e os impactos para as pessoas e para o negócio. Dessa forma, a IA deixa de ser uma exigência consequente tecnológica e passa a ser uma capacidade organizacional orientada à geração de valor, inovação, eficiência e consequentes vantagem competitiva, alinhada aos princípios dos padrões de projetos de IA do PMI... Saving Changes...
The first thing is to clarify what AI means. Human beings are using AI from more than 50 years ago. We are surrounded of AI entities embeded inside refrigerators, air conditioners, cell phones, etc, etc, Unfortunately in the last time some people and organizations are contributing to the general confusion using generative AI as a synonim of AI.
Hi there! Iagree. I also think it´s more important to understand what AI is, what it isn´t, and what limitations are. Unfortunalety, when people say, "we should use AI" they often imagine some incredible tool that can do everything automatically, without understanding what it actualiiy takes to make it work Saving Changes...
When someone says “we should use AI,” the conversation is rarely about technology itself. It is usually about pressure for speed, efficiency, innovation, or competitive leverage. The first step is to clarify intent.
Three signals help distinguish what is really being asked.
First, decision proximity. Is AI automating a task, augmenting human judgment, or moving toward managing objectives autonomously? These are fundamentally different categories of work. The closer AI gets to consequential decisions, the stronger the need for governance, traceability, and explicit oversight.
Second, problem clarity. Is there a clearly defined business problem with measurable impact, or is AI being treated as the starting point? When the solution precedes the problem, misalignment and inflated expectations follow.
Third, accountability design. Who owns the outcome if an AI-driven recommendation fails? When responsibility becomes diffuse, risk scales faster than performance.
In many organizations, “AI” simultaneously means efficiency, experimentation, and cost reduction to different stakeholders. Misalignment becomes visible when decision flows and ownership are unclear. A common tipping point is when stakeholders use the same word “AI” but describe different success metrics.
The real shift is not from manual to automated. It is from “man in the loop” to “man in control.” Without deliberate design of responsibility, capability increases while accountability erodes.
Clarity of purpose, category of AI work, and ownership separates disciplined transformation from technological noise.
In project management, for example, saying “we should use AI” is not enough. We need to identify the specific challenge, the expected outcome, the available data, and how success will be measured. The key is to start with the problem, not the technology.
I have a lot going on, and AI helps me include critical aspects of an overwhelming amount of information and guidance, ensuring that I don't forget to include anything. It also gives me a good starting point.
AI also assists after the initial build, by answering quick questions from leadership or stakeholder review. I'm able to defend my position more clearly with much less back and forth. It is a very natural tool to use in most parts of my projects to keep it on track Saving Changes...
When someone says, "We should use AI," I try to determine my response by the speaker. If it's one of my technical colleagues, I'm usually safe in assuming they think there is an AI tool that will complete a task more efficiently. If it's from management, they often mean they think we can use AI to find business insights that we are currently not capturing. Saving Changes...
style.ql-indent-1 { margin-left: 3em; }/styleWhen someone says, "We should use AI," it’s rarely a purely technical request. In my experience, it’s usually a reflection of a high-level goal, an operational frustration, or even executive pressure (FOMO). To translate that broad statement into a real, actionable strategy, I like to unpack it across four key dimensions:
Uncover the True Driver (The "Why"):
First, figure out what problem they are actually trying to solve. Are they looking to reduce manual work and cut costs? Are they trying to build a feature that wasn’t possible before? Or are they simply trying to align with market trends? Knowing the underlying motivation determines the metric for success.
Identify the Right Category of AI (The "What"):
"AI" is a massive umbrella term. We need to clarify if the problem calls for Generative AI (text/code synthesis, content creation), Predictive/ML models (data analysis, forecasting, risk assessment), or Smart Automation (document extraction, rule-based workflows). Matching the specific problem to the right subset of AI prevents misaligned expectations.
Assess Feasibility & Preparedness (The "Can We?"):
Before diving into execution, we have to evaluate our foundation:
Data Quality: Do we have structured, clean, and secure data to feed the model?
Governance & Risk: What is our tolerance for error or hallucination? Who stays accountable?
Human-in-the-Loop: How will human oversight validate the outputs, especially in critical workflows?
Ask the Right Diagnostic Questions:
To move from hype to scope, I always bring the conversation back to basics:
What specific bottleneck are we trying to eliminate?
Do we want AI to automate the process completely or augment our team as an assistant?
Could this be solved faster or more cost-effectively with a simpler, non-AI automation?
Ultimately, the goal isn't just to "use AI"—it's to solve a real business problem where AI happens to be the most efficient tool for the job. How do you usually handle these conversations when stakeholders bring up AI? Saving Changes...
Probably, the clearest signal is purpose: are we asking AI to predict, create, recommend, or decide? Those verbs reveal more than the label itself. When one person imagines a chatbot and another imagines an autonomous decision-maker, they may share a word without sharing a vision. Lumping everything together blurs expectations, risks, and responsibility. My starting question would be: “What human problem are we trying to solve? and what judgment should remain ours?” Clarity begins not with choosing a tool, but with naming the task. Saving Changes...
PMI Team Member
