Part one of our five-part series: Deploying Behavioural Assessments in Infrastructure Procurement
AI is already changing the quality and speed of tender production. As written submissions become easier to enhance, how do we distinguish between a compelling bid and the capability of the team expected to deliver it?
Recent* research reports that 79% of proposal teams have used generative AI in Request for Proposal responses, compared with 34% in 2023. UK construction research points in the same direction.
That development brings genuine benefits – AI can improve consistency, help teams navigate complex tender information and enable organisations to articulate good ideas more effectively.
The issue is not whether AI should be used, it is what the increasing sophistication of written submissions means for the evidence clients need alongside them.
From what a team can describe to what it can demonstrate
This is not a new procurement challenge. Behavioural and collaborative assessments have been used on major infrastructure programmes for many years. Long before generative AI, clients were looking for ways to test whether the people behind a tender could actually operate in the way the submission described. At that point, it becomes less about what is written on paper and more about understanding how the people responsible for delivery might actually work together.
AI does not create that need. It reinforces the importance of distinguishing between what a team can describe and what it can demonstrate. For example:
- A written response can explain how a team will manage uncertainty; a live assessment can see what happens when the information changes or established assumptions are challenged.
- A tender can describe collaborative decision-making; an assessment can observe how the nominated team navigates competing commercial, programme and stakeholder priorities.
- A bid can set out how challenge will be encouraged; an assessment can show what happens when people genuinely disagree.
These are different forms of evidence answering different questions. Technical and commercial evaluation tell us important things about the proposition. Past performance tells us something about previous delivery. Interviews can explore individual experience and judgement. Behavioural assessment can provide evidence of how nominated people interact, exercise judgement, make decisions together and demonstrate the behaviours required for successful delivery, particularly when faced with uncertainty or pressure.
What are we actually trying to evidence?
The question is broader than whether to include a behavioural assessment. There may be evidence we need at three different levels:
- Individual – do critical people have the judgement, experience and leadership capability needed for the role?
- Team – can these people work together, manage tensions and make effective decisions?
- Organisation – will the wider business provide the capacity, governance, resources and support needed to deliver the commitments being made?
Those questions may require different forms of evidence. A team exercise can reveal how nominated people interact. Interviews can explore individual judgement and experience. References, site visits, organisational evidence or clarification can help test whether the capability extends beyond the people in the assessment room.
Cabinet Office guidance on AI in procurement makes a similar point: use of AI in preparing a tender is legitimate, but buyers may still want proportionate assurance about the capability and capacity behind the submission.
The implication for procurement
The implication is not that AI means more behavioural assessment. It is that, as the written proposition becomes easier to strengthen, clients need to be increasingly deliberate about what capability needs to be demonstrated, by whom and through which forms of evidence.
ResoLex helps clients look beyond the quality of the bid to build confidence in the people, teams and organisations expected to deliver it. For bidders, we help close the gap between the proposition on paper and the capability the nominated team can demonstrate in practice. Our experience spans supporting leading engineering consultancies, contractors and programme-management organisations with behavioural assessment preparation on successful bids, as well as working with major UK infrastructure clients and programmes to develop and deliver behavioural assessment approaches.
Next up, we’ll move into what the delivery model requires and what needs to be evidenced: ‘Start with the delivery model, not the procurement route’.
If you want to learn more, contact Liz Lawman and Jay van den Boogaard: info@resolex.com
Source: * Loopio's 2026 and Rider Levett Bucknall's 2026 Procurement Trends