Project managers coordinate goals, people, decisions, dates and dependencies while the information around them keeps changing. AI can help organize that information into a useful first draft, while the project manager keeps ownership of scope, commitments and stakeholder decisions.

A useful starting point is a task that happens regularly and already has a clear quality bar. Examples include turning a meeting transcript into a decision log, preparing a weekly update from team notes or structuring a risk review. The team knows what a useful result contains, so it can check the draft against its definition of a complete update.

Start with the project routine

Before choosing a tool, map the routine. What information arrives? Who contributes? Which judgment does the project manager apply? What does the next person need? This short map shows where a draft, summary or comparison could save attention.

For example, a project update may pull together progress notes, decisions and blockers from several workstream owners. An AI step can group the material by workstream, flag dates that need confirmation and prepare a draft in the company’s agreed format. The project manager reviews the source notes, confirms changes to commitments and sends the update with clear ownership.

The workflow can be written in five steps:

Team notes and approved project material
    ↓
AI groups facts, actions and open decisions
    ↓
Project manager checks sources and dates
    ↓
Owners confirm their actions and commitments
    ↓
Shared status update and next review date

This pattern works for many project routines. AI structures information; the project manager interprets it in context and leads the conversation about what happens next.

Use AI for work people can review

Meeting follow-up is a useful first scenario. A project manager can ask AI to draft action items from notes, group questions by decision owner and produce a concise recap. The review checks names, deadlines, dependencies and whether each item represents a decision or a proposal.

Planning can use the same idea. Give AI a defined objective, a set of confirmed constraints and a sample project brief. Ask for a draft work breakdown, a list of dependencies or alternative ways to sequence activities. The project manager compares the suggestions with team capacity, delivery commitments and organizational priorities.

Risk review benefits from structured thinking as well. AI can group risks already identified by the team, prepare questions for a workshop and draft possible signals to monitor. A person familiar with the project confirms likelihood, impact, owner and response. These steps create a stronger discussion while keeping risk decisions with the responsible team.

PMI’s 2025 report, The Project Professional’s GenAI Journey, describes a relationship between experimentation and reported productivity gains. In its comparison, 48% of “Trailblazers” and 3% of “Explorers” reported advanced productivity gains. The groups in that study reflect different levels of AI practice; a project team can use the finding as a reason to practise deliberately and measure its own baseline. (PMI, The Project Professional’s GenAI Journey)

Set a clear review standard

Useful AI practice gives each output a named reviewer and a short checklist. A project status draft might be checked for source, date, owner, confidence and impact on the plan. A decision summary might be checked for the decision-maker, options considered and the next action.

The Standard for Artificial Intelligence in Portfolio, Program and Project Management from PMI includes human-in-the-loop practices, governance, risk and applied project use cases. It reflects a practical principle for project work: AI output belongs inside an accountable workflow, with roles and review criteria defined alongside the task. (PMI, The Standard for Artificial Intelligence in Portfolio, Program and Project Management)

For AI-driven initiatives, PMI’s Leading & Managing AI Projects guide describes a lifecycle approach that connects organizational goals, data readiness, governance and ongoing measurement. Project managers can use these ideas even when the team’s first experiment is small: define a goal, check the available information, assign a reviewer and evaluate what changed. (PMI, Leading & Managing AI Projects Digital Guide)

Choose the right learning pace

A one-day lab maps one familiar task, tests an AI step with approved sample material and produces a review checklist. A one-week workflow sprint develops one recurring project workflow, such as a weekly status report, into a documented test with an owner and a next-step decision. A four-week program brings several roles together around workflow cards, a shared review standard and a priority map for the next pilot.

The outputs make each format easy to compare: one task and test for the day, one process and decision for the week, and a set of role-based practices and shared priorities for four weeks. A project lead can pair the learning record with a task measure, such as preparation time, missing owners or dates, or the number of corrections requested in the next work cycle.

Make the experiment part of delivery

Treat AI practice like a small project experiment. Record the current process, choose a specific output to improve, agree on review criteria and test the workflow with approved example material. After the test, compare the draft with the team’s standard and capture the corrections that made it useful. For example, track the time needed to prepare a status and count how often a reviewer returns it for missing dates, owners or dependencies.