What PMI's 2026 Standard for AI in Project Management Means for Us
- alisabolander
- Aug 26
- 2 min read
PMI’s new Standard for Artificial Intelligence in Portfolio, Program and Project Management (June 2026) has me thinking about a particular risk of generative AI:
Gen AI can turn epistemological uncertainty into apparent (and misleading) linguistic certainty.
Give a Gen AI model incomplete context and it may still produce a polished charter, plan, analysis, or recommendation. The output can look far more resolved than the underlying information actually is.
Instead of a blank project charter template, we’ll have something that appears filled out, complete, and ready to action. There will be objectives, WBS, tasks, and targets. There will be general names for general types of activities that can and do exist in certain types of projects. The agenda will list general flow for general types of topics to discuss, generally.
PMI noticed this creates a new kind of quality and governance question for project teams: Before an AI-generated artifact enters a workflow, do we know which parts came from validated inputs and decisions, which are assumptions, and which still require a person to define?
Human-in-the-loop shouldn’t just mean reviewing the final output or artifact.
Sometimes the most important human judgment happens before AI generation begins: defining the business problem and business case, supplying the best available relevant context, requests for identifying what isn’t yet known and what questions have not yet been asked, and deciding what the AI model is actually being tasked to do as it holds a microphone up to amplify the quality and specificity of the human-supplied prompts and inputs.
That’s one of my biggest takeaways as I dig into PMI’s new AI standard for us. What has PMI’s guidance got you thinking about differently this summer?
Comments