The contract is now auditable.

The gap between the duty and the practice

A significant part of the industry runs commercial bias as standard. It is treated as an implicit part of the service. The firm is appointed by the employer, paid by the employer, and answers to the employer, so the administration is shaded in the employer's favour. Early warnings get suppressed when they expose the employer to a change. Time and cost assessments get framed in a way that minimises the employer's exposure. Impartial decisions get replaced with negotiated ones. Correspondence gets drafted with one eye on legal defensibility rather than on the contract's actual requirements.

The bias rarely shows up in any single dramatic moment. It shows up in the pattern of small decisions taken across a year. The early warning that should have been issued in February that was not, because issuing it would have triggered a compensation event the employer wanted to avoid. The programme acceptance withheld in May for reasons that would not survive scrutiny if anyone scrutinised them. The change event in July assessed in a way that pushed the cost back onto the contractor rather than where it actually belonged. Each one looks reasonable in isolation. The pattern across the project tells a different story.

"The bias is not in any single letter. It is in the pattern across twelve months of correspondence."

What the law actually says

The legal position has been clear for half a century. Sutcliffe v Thackrah, decided in 1974, is the foundational case. Lord Reid's judgement is the one every contract administrator should know. The employer and the contractor make their contract on the understanding that in all matters where the administrator has to apply professional skill, they will act in a fair and unbiased manner in applying the terms of the contract.

Costain v Bechtel, decided in 2005, extended the same principle to the New Engineering Contract. Hickman v Roberts, dating from 1913, is the oldest authority and still the sharpest. The architect openly admitted to the contractor that the employer would not allow him to issue a certificate. The certificates were set aside. That was the law a century ago. It is the law today.

What sits between the law and the practice is cost. The duty has always been there. The mechanism for proving a breach has been prohibitively expensive. A client who suspected their administrator was running bias has historically had very limited options. The only route was a forensic review of the correspondence and the contract record by a third party. Slow. Expensive. Usually only initiated after something had gone badly wrong. Most clients, most of the time, paid for the bias without ever proving it, because proving it cost more than absorbing it.

What has changed

A client today can take the contract, the programme updates, the change registers, the correspondence record and the cost reports, and put them through a model that will return a structured analysis in minutes. Where was the early warning the contract required. Where was the impartial decision that went the employer's way. Where did the administrator's advice serve the firm's position rather than the contract's terms. Where did the aggressive letter replace the decision the contract actually asked for.

None of these questions are new. What is new is that they can be asked cheaply, quickly, and repeatedly, by someone who does not need forty years of construction law experience to frame them.

"The questions used to require weeks of forensic work. They now take minutes."

The patterns the analysis returns are the same patterns the experienced eye sees, but laid out across the whole correspondence record at once, with the gaps named and the inconsistencies tagged. The early warnings that should have been issued and were not. The compensation event quotations assessed against criteria that drift through the year. The certifications that do not match the programme evidence. The decisions that consistently land on one side of the line where the contract requires balance. None of it is invisible. It is just visible in a way that used to require weeks of forensic work and now takes minutes.

What this means for the firms

The tactics the traditional model uses to manage exposure are the ones most exposed by the new transparency. The letter-writing campaign in month eleven. The contractual argument about who bears a risk that should have been spotted in month three. The inflated tender return defended as the market. The claim framed around legal defensibility rather than factual accuracy. All of it looked like professional sharpness when the only audience was the client and the only reading was a human one. It looks very different when the audience is a model that has read the contract and has no stake in the firm's reputation.

The aggression is not incidental to the traditional model. It is part of how the model holds together when its own decisions need defending. As long as the audit cost stayed high, the defence worked. Once the audit cost collapses, the same correspondence stops being a defence and starts being a record.

"What looked like defensive sharpness becomes documentary evidence of a firm covering its earlier errors."

What this means for clients

Choosing a consultancy used to be a reputational exercise. Clients picked the firm with the history, the letterhead, the roster of past projects, and the partners who looked the part. The risk of getting that decision wrong was limited, because the evidence of whether the firm was any good was locked inside processes that were almost never audited. That risk is no longer limited. A client who appoints a consultancy on reputation alone, without looking at how that consultancy's administration actually performs under the contract, is taking on an audit risk that did not exist five years ago.

The clients who are leading on this are already asking different questions when they interview firms. They want to see sample correspondence from past projects. They ask about early warning discipline. They ask how change events are assessed. They ask whether the firm treats the contract administrator's impartiality as a working principle or a legal inconvenience. They are doing, in their pre-appointment process, the kind of audit that used to happen only after a project had gone badly wrong, because the only reliable way to avoid paying for a decade of defensive correspondence is to not appoint the firm that writes it.

There is a quieter point in all of this. A contract administered impartially for its full term, with a clean correspondence record that reads true under audit, is a more valuable asset to both parties than a contract administered to protect one side's position. Impartial administration is not a concession to the contractor. It is an enhancement of the contract's value to the employer, because the resulting project finishes better, the final account settles rather than litigates, and the supply chain takes the next call.

The light has been turned on. The firms whose model was built to stand up to the visibility will be fine. The ones whose model depends on the visibility staying off will not be, and the timeline is not long.