Senior project leaders share practical insights on AI in project delivery, from governance and data integrity to the risk of false confidence. Their perspectives highlight emerging opportunities for the PMO and where skilled human oversight remains essential.
Key Insights
- False confidence is a major risk
- Data governance and security matter — without clear ownership and enterprise tools, organisations risk data exposure and poor quality.
- Human expertise turns AI into value
AI is making its way into project delivery at every level, from high-value programs to informal use in smaller, business-led initiatives.
While adoption is in its early days, the potential for what we call Amplified Impact — the tendency for AI to magnify whatever delivery practices are already in place — is already becoming evident.
In mature environments with strong governance, it could help accelerate delivery and support better decision-making. Without that foundation, it risks creating a sense of progress that may not reflect the reality of delivery. For a deeper look at the concept, see our explainer article on Amplified Impact in project delivery.
To explore how this potential is starting to emerge in practice, we spoke with senior practitioners and advisors from across the Quay Consulting network. Their perspectives span technology governance, delivery assurance, organisational change and the hands-on reality of running projects. Together, they offered a grounded view of where AI might create value, where it could introduce new exposure and how project delivery and how the PMO can remain central to safe, effective adoption.
Our Expert Panel
Executive advisor and recognised authority on complex program delivery and strategy execution, with extensive experience in independent assurance for high-stakes initiatives across industries. |
International speaker, educator and consultant on project leadership, influence and communication, with over three decades of strategic consulting and systems delivery across Africa, Europe and the Asia Pacific. |
PMO and program manager with extensive experience in business transformation, ERP delivery and strategic planning across sectors including construction, government, manufacturing and not-for-profit. |
|
Product and transformation leader with experience in education, financial services, health and FMCG, specialising in organisational change, technology adoption and scalable product delivery. |
Practice Director at Quay Consulting and transformation leader with extensive experience in program delivery, portfolio governance and building high-performing teams for complex business and technology initiatives. |
|
False confidence and the DIY risk
Several of our experts pointed to a common concern: AI can make project work look more complete than it really is, creating a false sense of confidence. The danger grows when teams lack the delivery maturity to recognise what is missing.
Mike Kaye: “I am seeing clients treat AI as a way to DIY parts of a project — particularly with things like schedules, risk registers and reports. On the surface it looks fine, but when you dig into the detail there is not much there. If you do not know what good looks like, you will not see the gaps.”
Jon Pascoe: “I asked an AI tool to review a project variation by looking at the business case, project plan and governance packs. It came back with dates and facts, but it completely missed the most important point — there was no clear evidence of forward progress. The detail was there, but the substance was not.”
Derek Keogh: “It is not just about asking AI to produce a report. You have to know what to ask next. For example, you might ask it to compare an output to a well-run project and tell me what is missing. That layer of thinking is what turns a basic output into something useful.”
For decentralised or business-led projects working with tight budgets, the risk presented by overconfidence is even higher. AI can give the impression that the work is on track when the fundamentals are not in place. The PMO’s role in setting quality standards and providing independent oversight is essential to keep incomplete work from moving forward unchecked.
To find out more about the dynamics of these lean, business-led projects, see our recent article on the role of fractional support in de-risking business-led delivery.
Project governance and data integrity
As organisations start to experiment with AI in project delivery, questions of governance and data integrity are coming to the forefront. Without clear rules on ownership, storage and use, the risks extend beyond individual projects to the organisation as a whole.
Jürgen Oschadleus: “When you talk about AI in the PMO, governance is not just about the tool itself. It is about who owns the data, where it is stored and how it is maintained for security and integrity. If those fundamentals are weak, you are creating exposure.”
Abby Clifton: “There is a real split between people using personal tools like ChatGPT and those using enterprise tools such as Copilot. Without governance around that, you end up with valuable business data sitting outside secure environments — and often without the organisation even knowing.”
Jon Pascoe: “Another factor is where the AI is located. Major platforms are rapidly bringing AI capabilities on-platform, rather than relying on third-party services. As that happens, the opportunities grow exponentially, because the data on-platform can be fully interrogated and applied to AI-enabled use cases.”
Jürgen Oschadleus: “In the longer term, PMOs will need more people with data analysis skills rather than traditional project analysts. You need that capability early, so you can set up the data in a way that makes it useful for AI tools without compromising its quality or security.”
Effective governance supports safe adoption by ensuring the right conditions are in place from the outset. For the PMO, this involves setting clear policies, building the right skills and maintaining oversight so that AI outputs draw on accurate and well-managed information.
The value of human expertise
While AI can help with speed and efficiency, it cannot replicate the judgement and contextual understanding that experienced practitioners bring to delivery.
Derek Keogh: “You get the best results when you treat AI as an assistant, not an autonomous operator. It’s the follow-up questions, the comparisons, the gap analysis that turn an output into something useful for decision-making.”
Human expertise is what turns AI into a productive tool rather than a source of false assurance. For the PMO, the focus should be on using technology to augment skilled practitioners, not replace them.
Opportunities for the PMO
AI is already demonstrating value in specific, well-defined use cases. For the PMO, there is an opportunity to guide adoption so these early wins translate into sustainable delivery benefits.
Abby Clifton: “Where AI has a defined use case for project delivery — whether it is meeting notes, reporting or a targeted process improvement — and the right governance is in place, the productivity gains can be significant. The key is getting teams to collaborate around those use cases rather than working in isolation with personal tools.”
By taking the lead on defining safe, value-focused applications, the PMO can help organisations integrate AI in ways that enhance delivery capability and maintain control over the quality of outcomes. Read more about this in our deep dive into the AI use cases we see as most relevant for project delivery, and how PMOs can prioritise them to create measurable impact.
What good looks like in AI-enabled project delivery
Overall, our experts pointed to a future where AI is applied in project delivery to serve well-defined priorities. A shared understanding of value will help guide decisions about how and where to use the technology.
Governance should enable progress while still protecting against unnecessary risk, and practitioners will need to combine knowledge of the delivery context with an informed approach to the tools in use. Reliable, well-managed data will be essential, but the real differentiator will be a culture that prioritises enquiry — testing and refining outputs, rather than accepting them at face value.
If you’re exploring where AI might fit into your project delivery approach, considering how to build the right foundations, or want to test whether your governance can support safe adoption, our experts are always up for a conversation. Let’s brainstorm it together.
Quay Consulting is a professional services business specialising in the project landscape, transforming strategy into fit-for-purpose delivery. Meet our team or reach out to have a discussion today.
Please share if you’ve found value in our content. If you’re interested in republishing our articles, please, check out our guidelines.
