AI in Project Management: Who’s actually governing it?

AI is already changing project management.

Project professionals are using AI to summarise meetings, draft reports, analyse information, generate ideas, identify risks and reduce hours of project administration.

And that’s exciting.

But there’s a question getting far less attention:

Who’s actually governing it?

Because using AI to save 30 minutes on a project update is one thing.

Allowing AI-generated information to influence a major project decision is something else entirely.

As AI becomes part of everyday project delivery, Australian project managers may need to develop a new capability alongside planning, risk management and stakeholder engagement:

AI governance.

AI in project management has moved beyond experimentation

It wasn’t long ago that using generative AI at work felt experimental.

That’s changing quickly.

AI is increasingly embedded within the tools people already use, which means project teams don’t necessarily have to make a conscious decision to “adopt AI”.

It can simply appear in their workflow.

Think about the possibilities.

A project manager asks AI to summarise a lengthy document.

A PMO uses it to help identify trends across project data.

A team asks an AI assistant to draft a risk description.

Someone uses generative AI to prepare a stakeholder communication.

Another employee pastes information from a project into a public AI tool because they need a quick answer.

Each individual action might seem relatively harmless.

Collectively, they raise a much bigger question.

Do you know how AI is being used across your projects?

The project manager’s new risk: Shadow AI

This is where Shadow AI enters the picture.

Shadow AI broadly describes employees using AI tools or applications without appropriate organisational approval, visibility or governance.

And it’s easy to understand why it happens.

Imagine you’re under pressure to finish a project report.

An AI tool can turn your notes into a polished first draft in seconds.

Would you use it?

Many people would.

The problem isn’t necessarily the employee’s intention.

The problem is what happens when organisations adopt AI faster than they establish the controls needed to manage it.

What information was entered into the AI?

Where did that data go?

Was the tool approved?

Was its output checked?

Could the answer contain inaccurate information?

Who is accountable if that information influences a project decision?

Suddenly, a productivity tool becomes a project governance issue.

What happens when AI confidently gets it wrong?

Anyone who has spent time using generative AI knows one of its most interesting characteristics:

It can sound extremely convincing even when it’s wrong.

For project professionals, that’s an important risk.

Imagine an AI tool analysing a project schedule and identifying what it believes is the biggest threat to delivery.

The answer looks logical.

The report is well written.

The recommendation sounds convincing.

But the model has misunderstood part of the project data.

If nobody challenges the output, that mistake could influence resources, priorities or decisions.

The problem isn’t simply that AI can make mistakes.

People make mistakes too.

The governance question is:

What controls exist to catch them?

Australia is putting greater emphasis on responsible AI

AI governance isn’t just a theoretical concern.

Australia’s approach to responsible AI is developing as adoption increases.

The Australian Government’s Policy for the responsible use of AI in government, updated in December 2025, includes requirements covering AI strategy and oversight, accountability, internal AI use-case registers, staff training and AI use-case impact assessments.

The policy also emphasises something particularly relevant to project professionals:

People using AI need to be able to explain, justify and take ownership of their advice and decisions.

That’s an important principle whether you work in government or not.

AI can support a decision.

It doesn’t remove human accountability for making it.

Five AI governance questions every project manager should ask

You don’t need to become an AI engineer to start governing AI more effectively.

Project managers can begin with five practical questions.

1. Where is AI being used?

Start with visibility.

Which AI tools are officially approved?

More importantly, which tools are people actually using?

AI could already be appearing in reporting, scheduling, risk management, stakeholder communications, analysis or decision support without being formally recognised as part of the project environment.

You can’t govern something you don’t know exists.

2. What information are we giving AI?

Consider the data being entered into AI systems.

Does it include commercially sensitive information?

Customer or employee data?

Financial information?

Confidential project documentation?

Before using an AI tool, project teams need to understand how information is being handled and whether its use is appropriate.

3. How are we checking AI-generated outputs?

AI should not automatically become the project’s source of truth.

Teams need to understand when outputs require human review and who is responsible for that verification.

The higher the potential impact of an error, the stronger that oversight may need to be.

4. Who owns the decision?

This might be the most important question of all.

If AI recommends a course of action, who decides whether to follow it?

And who is accountable for the outcome?

“AI told us to do it” isn’t a governance model.

Projects still need clear human accountability.

5. What happens when something goes wrong?

Good governance doesn’t assume everything will work perfectly.

What happens if AI exposes sensitive information?

Produces misleading analysis?

Creates biased output?

Influences a poor decision?

Project teams need appropriate escalation and response processes just as they would for other significant project risks.

AI governance isn’t about saying “no” to AI

The word governance can make people nervous.

It sounds like rules.

Approvals.

Restrictions.

More paperwork.

But good AI governance shouldn’t be about stopping people from using AI.

It should make it easier to use AI with confidence.

Think about risk management in a project.

We don’t identify risks because we want everyone to stop doing anything remotely uncertain.

We identify them so we can make informed decisions.

AI governance should work in much the same way.

The objective is to gain the productivity and innovation benefits of AI while understanding and managing the risks that come with it.

Why project managers are well placed to lead AI governance

Here’s where this becomes particularly interesting for the project profession.

Project managers already know how to govern change.

They think about:

• Roles and responsibilities
• Risk
• Controls
• Stakeholders
• Escalation
• Decision-making
• Benefits
• Governance
• Organisational readiness

Those skills are highly relevant to AI adoption.

That means the AI revolution doesn’t necessarily require every project manager to become a technical AI expert.

Instead, organisations need people who can bridge the gap between what AI can do and how it should be used responsibly within a project environment.

That could become an increasingly valuable project management skill.

What is the AI Project Governance Framework?

The AI Project Governance Framework (AIPGF) is a methodology developed specifically to help project professionals and oversight roles integrate AI governance into existing projects and programmes.

It’s designed to be scalable according to factors including project size, complexity, risk and an organisation’s level of AI maturity.

Importantly, it isn’t designed to replace the project management approach you’re already using.

AIPGF is methodology-neutral.

That means it can provide an AI governance layer alongside established approaches such as PRINCE2, APM and Agile.

And it isn’t a course in coding or building artificial intelligence.

The focus is much more relevant to project professionals:

Should we use AI here?

What risks do we need to consider?

Who should be accountable?

How do we govern its use appropriately?

Is AI governance becoming a project management skill?

Project management has always evolved alongside the way organisations work.

Digital transformation changed it.

Agile changed it.

Remote and hybrid working changed it.

AI will change it again.

But perhaps the biggest opportunity for project professionals isn’t learning how to generate better prompts.

It’s learning how to make sure AI is used ethically, effectively and responsibly within the projects they lead.

The Australian Government’s current approach provides an interesting signal. Its responsible AI policy now requires affected agencies to establish clear accountability for AI use cases and introduces risk-based oversight and impact assessment requirements.

As AI adoption grows elsewhere, organisations are likely to need people capable of turning principles like these into practical governance.

Project professionals are already well positioned to do exactly that.

The future isn’t human vs AI

The future of project management probably isn’t about choosing between human judgement and artificial intelligence.

It’s about working out how the two work together.

AI can analyse.

AI can automate.

AI can generate.

AI can recommend.

But people still need to question, interpret, govern and take responsibility.

That’s where project managers come in.

So, if AI is already appearing in your projects, perhaps the question isn’t:

“Should we be using AI?”

It’s:

“Are we governing it properly?”

Learn how to govern AI in projects

Training ByteSize Australia offers the APMG AI Project Governance Framework (AIPGF) Certification, designed for professionals who need to understand and manage the governance implications of AI within projects and programmes.

The training covers areas including AI ethics and law, roles and responsibilities, organisational AI maturity, governance, tailoring and the ethical, efficient and effective use of AI.

It’s particularly relevant to project, programme and portfolio managers, PMO leaders, senior stakeholders, change managers and professionals already working with approaches such as PRINCE2, APM and Agile.

Rather than teaching you how to code AI, AIPGF helps answer the project management questions surrounding it.

Because organisations don’t just need people who know how to use AI.

They increasingly need people who know how to govern it.

Explore the APMG AI Project Governance Framework (AIPGF) Certification with Training ByteSize Australia.