AI in Project Management: What Actually Helps a Technical PM

AI is everywhere in project management conversations right now, but using it well is still rare. In a September 2025 report, PMI described generative AI adoption as skyrocketing while organisational maturity is still hovering around just 1%, mostly because organisations lack the processes, leadership alignment, and readiness to use it properly.

I'm a technical project manager, I still write code, and I completed PMI's Generative AI Overview for Project Managers course. This article is my practical view of AI in project management: where it helps, where it doesn't, and how to start without creating new risks.

The short answer

AI is good at drafting, summarising, structuring, and checking. It is not good at judgement, accountability, and relationships. Use it to remove the repetitive parts of the job, so you have more time for the parts only a person can do: aligning people, making trade-offs, and taking responsibility for outcomes.

Where AI genuinely helps

Turning messy input into structure

A lot of project work starts as unstructured text: meeting notes, long email threads, a client's feature list in a single paragraph. AI is good at turning that into something usable, for example:

• a first draft of user stories with acceptance criteria,
• a list of open questions to clarify with the client,
• a summary of decisions and action items from a meeting transcript.

The output always needs review, but starting from a structured draft is much faster than starting from a blank page.

First drafts of documents

Status reports, release notes, risk descriptions, change-request summaries, and onboarding notes follow predictable patterns. AI can produce a solid first draft in seconds, and your job becomes checking accuracy and tone.

Planning and risk support

AI can help break a feature into tasks, suggest dependencies you might have missed, or list the risks typically associated with a type of work, such as a payment integration or a data migration. Treat the result as a checklist to think through, not as a plan. PMI's 2024 research on GenAI adoption found that the heaviest users apply it to more complex work like planning, data analysis, and risk management, not only to writing.

Testing and quality

This is where my engineering background matters most. AI-assisted tools can help generate test cases from requirements, suggest edge cases, and support automated browser and API testing. Used well, they help QA cover more ground in the same time.

Automating repetitive work

Many small, recurring tasks can be automated: collecting status from a tracker, formatting a weekly summary, or flagging tasks that have been blocked too long. I automate as much as I can, because every repetitive task that disappears gives time back for real problems.

AI tools and agents in project management

The biggest recent change is that AI is moving inside project management tools. In 2026, Atlassian made it possible to assign Jira work items to AI agents, @mention them in comments, and add them to workflows, with admins controlling which agents are allowed and every agent action logged. Other tools are moving in the same direction.

This is useful, but the same rule applies as with any team member: an agent needs a clear task, a defined "done", and someone accountable for reviewing its work.

Where AI does not help (yet)

• Judgement calls. Deciding whether to cut scope, add a resource, or push back on a client depends on context, relationships, and trade-offs that aren't in any document.
• Accountability. If an estimate is wrong or a risk is missed, "the AI suggested it" is not an answer. The project manager owns the outcome.
• Trust and relationships. Clients and teams trust people who listen, follow up, and keep their word.
• Facts it doesn't have. AI tools can produce confident answers that are simply wrong. Anything factual (dates, numbers, technical claims) must be checked.

The risks to manage

Using AI in a delivery team creates a few new responsibilities:

1. Confidentiality. Never paste client data, credentials, contracts, or internal details into tools that aren't approved for that use. Check your organisation's policy and the tool's data settings first.
2. Accuracy. Review every output before it reaches a client or the team.
3. Over-reliance. If the team stops thinking because the tool always suggests something, quality drops quietly.
4. Transparency and governance. Be clear about where AI is used in your process and who reviews it. In June 2026 PMI published The Standard for Artificial Intelligence in Portfolio, Program and Project Management, the first ANSI-approved AI standard for the profession, which focuses on exactly this: moving from ad hoc AI use to structured, accountable practice.

How to use AI in project management, step by step

1. Pick one repetitive, low-risk task that costs you time every week, such as meeting summaries or the first draft of a status report.
2. Use an approved tool and remove anything confidential from your input.
3. Always review the output against what you actually know.
4. Measure the result: did it save time, and was the quality the same or better?
5. Expand slowly to the next task, and share what works with the team.

What PMI's free GenAI course covers

PMI's Generative AI Overview for Project Managers is free, takes about an hour, and earns 5 PDUs. It's split into five modules:

1. Introduction to GenAI: what it is, how it can improve project outcomes, and what to consider before using it.
2. Enhancing PM with GenAI: applying it across PMI's Talent Triangle (ways of working, business acumen, and power skills).
3. Voice of the PM: project managers describing how they use GenAI for scheduling, planning, documentation, risk logging, and estimating.
4. ChatGPT Lab: prompting techniques, with a downloadable prompt engineering guide.
5. AI Tool Library: a curated set of tools to try, including PMI's own AI assistant, PMI Infinity.

It's an introduction rather than a deep dive, which makes it a sensible first step. PMI has since added more free courses, such as Practical Application of Generative AI for Project Managers, and a certification for leading AI projects, PMI-CPMAI, for those who want to go further.

Frequently asked questions

Will AI replace project managers?

AI is taking over parts of the work, such as drafting, summarising, and routine tracking, but project management depends on judgement, trade-offs, alignment, and accountability. Those stay with people. The realistic change is that project managers who use AI well will spend less time on admin and more on decisions.

What are the best AI tools for project management?

The most useful AI tools are usually the ones built into what your team already uses, such as AI features and agents in your tracker, plus a general assistant approved by your organisation. Start with one tool and one task, and judge it by time saved and quality, not by features.

Is it safe to use AI tools with client data?

Only if your organisation has approved the tool for that kind of data and its settings protect it, for example by not using your inputs for training. When in doubt, remove client names, credentials, contract details, and anything confidential before using any AI tool, and agree with the client how AI is used on their project.

Is PMI's Generative AI Overview for Project Managers free?

Yes. PMI lists it as free, about one hour long, and worth 5 PDUs, which count towards renewing PMI certifications. It covers GenAI basics, applying it across PMI's Talent Triangle, real examples from project managers, prompting techniques, and a library of AI tools to try.

Final thoughts

AI won't run your projects. Used carefully, it gives you back the hours spent on repetitive work, so you can spend them on what actually moves a project forward: people, decisions, and follow-up.

Related: How I Track 10 Projects at Once and my journey from developer to project manager.