Artificial intelligence is already changing the way project managers work. For many people, AI still means ChatGPT, Copilot or Claude. They see it as a useful tool for drafting emails, summarising documents or finding quick answers.
But that is only the first stage.
In a recent Parallel Project Training podcast, Paul Naybour spoke with James Gardner from Project Flux about how AI is being used in project delivery today, and where it may go next. The discussion explored three levels of AI adoption in project management: knowledge assistance, decision support and organisational learning.
Level 1: AI as a knowledge assistant
The most common use of AI today is as a knowledge assistant.
At this level, AI helps people search, interpret and apply information. This could include project management textbooks, organisational procedures, contract guidance, standards, templates or lessons learned reports.
Many organisations have extensive libraries of guidance, but people often struggle to find the right information at the right time. AI can make this easier by acting as an intelligent search and explanation tool.
For example, a project manager might ask:
“What does our change control procedure say about approving a major scope change?”
Or:
“What guidance does the APM Body of Knowledge give on stakeholder engagement?”
This is particularly useful where the body of knowledge is constrained and reliable. If the AI is directed to use a specific set of documents, it can provide more focused answers and reduce the risk of hallucination.
However, this level of AI use requires a different mindset from traditional search. We have spent years learning to use Google by typing short keyword searches. AI works better when given context. The more clearly we explain the task, the source material and the intended output, the better the answer is likely to be.
Project professionals therefore need to develop AI literacy. They need to understand how to ask better questions, how to check sources, and how to use AI outputs responsibly.
Level 2: AI as decision support
The second level is decision support.
This is where AI moves beyond answering questions and starts to connect with project data. It can draw on schedules, cost systems, risk registers, change logs, dashboards, drawings, emails and meeting notes.
This is a major step forward.
Rather than asking the PMO to spend two days building a spreadsheet or pivot table, a project director could ask:
“What are the top ten risks across the portfolio?”
“Which changes are running late?”
“Which projects are showing early signs of cost pressure?”
“Summarise the key issues for my meeting with the contractor next week.”
This does not remove the need for professional judgement. The project manager still has to interpret the information and decide what action to take. But AI can remove much of the manual effort involved in gathering, sorting and summarising data.
This is also where AI agents and tools such as Model Context Protocol, or MCP, become important. MCP allows AI systems to connect more easily with business software. It means AI can work inside project systems, rather than simply talk about them.
For project managers, this could mean practical support with meeting preparation, contract reviews, risk analysis, stakeholder briefings and scenario planning.
AI can also act as a thought partner. For example, a project manager preparing for a difficult conversation with a contractor could ask AI to role-play the likely arguments, test possible responses and identify weaknesses in their position.
This use of AI is still emerging, but it has significant potential.
Level 3: AI for organisational learning
The third level is perhaps the most interesting. This is AI used for organisational learning.
Projects generate huge amounts of data. Some of it is numerical, such as cost, schedule, risk and performance information. Much of it is written, including emails, reports, meeting minutes, change requests, claims, lessons learned and stakeholder communications.
As humans, we are not very good at seeing patterns across large volumes of project data. We also struggle with lessons learned. They are often written at the end of the project, filed away, and rarely used effectively on the next project.
AI offers a different possibility.
It can analyse large volumes of project information and look for patterns, themes and changes over time. It can help identify where projects started to go wrong, what people were concerned about, where stakeholder sentiment changed, and which issues appeared repeatedly across a portfolio.
This could help organisations move from traditional lessons learned to live learning.
Instead of waiting until the end of a project, AI could support continuous reflection. It could identify early warning signs, compare projects with similar characteristics and help teams learn while there is still time to act.
This has important implications for portfolio management, governance and project assurance. It could help organisations understand not just whether projects are late or over budget, but why those patterns keep recurring.
The risks of AI in project management
The podcast also explored the risks.
AI is moving very quickly. The tools are becoming more powerful, more connected and more agentic. That creates opportunities, but it also creates governance challenges.
Project organisations need to think carefully about data security, privacy, accountability, hallucinations, intellectual property, bias and professional responsibility.
There is also a danger that people treat AI outputs as authoritative without proper scrutiny. AI can support judgement, but it should not replace it. Project managers still need to understand the context, challenge the evidence and remain accountable for decisions.
This is particularly important in areas such as contract management, risk, procurement, assurance and stakeholder communication.
Responsible AI use is therefore not just a technical issue. It is a professional competence issue.
What project managers should do now
AI will not remove the need for project managers. Projects still depend on people, judgement, negotiation, leadership and emotional intelligence.
However, AI will change what project managers spend their time doing.
Routine administration, document searching, meeting preparation and data analysis are likely to become much faster. This should free project professionals to spend more time on judgement, stakeholder engagement, decision-making and leadership.
Project managers should therefore start developing their AI literacy now. That means learning how the tools work, experimenting with practical use cases, understanding the risks, and thinking about how AI can support better project outcomes.
The starting point does not need to be complicated. Use AI to summarise a meeting. Ask it to explain a procedure. Use it to prepare for a difficult conversation. Test it against a known body of knowledge. Then gradually move towards more advanced uses, such as decision support and portfolio analysis.
The organisations that benefit most from AI will not simply be those with the newest tools. They will be the organisations that combine AI with strong project discipline, reliable data, professional judgement and a clear approach to governance.
AI is not a magic answer to project failure. But used well, it could help project managers make better decisions, learn faster and deliver more successful projects.
Listen to the podcast
In this episode, Paul Naybour talks to James Gardner from Project Flux about the practical use of AI in project management today and how it may shape the future of project delivery.
They discuss knowledge assistants, decision support, AI agents, MCP, organisational learning, lessons learned, portfolio insight and the risks of moving too quickly without proper governance.
