FIDIC CONTRACTS- HUMAN BEHAVIOURS, AI AND CARBON MANAGEMENT

This article contains links which are only accessible by PLC subscribers.

The FIDIC International Contract Users' Conference 2025 that took place in December 2025 in London opened with a humbling discussion for contract drafters and well-intentioned lawyers that centred on a key reality: contracts don't deliver projects; people do.

This candid reminder (and only a marginally dispiriting thought for someone having built a career crafting these contracts) continued throughout numerous sessions. It may seem that FIDIC were aiming to modestly deflect the importance of the contracts we were all gathered to analyse and discuss, however I think it was a refreshingly honest and practical approach. For the organisation (and us drafters) to reflect upon why contracts fail and what behaviours we can employ, and indeed encourage, through contract drafting and procurement strategy, was innovative in itself.

AI, the Engineer's accountability, cultural impact and human behaviour (including effective contract management), all themes explored at this year's conference, are shaping the future of FIDIC contracts and how projects are delivered.

For more information about FIDIC contracts generally, see Practice note, FIDIC Forms of Contract.

The Engineer – can they truly be "neutral"?

In FIDIC's popular red book (RB) and yellow book (YB) contracts, the Engineer is central to the contract's administration, performing a dual role of contract administrator (and certifier of payments), as well as making determinations under sub-clause 3.7. A key topic of discussion was whether the Engineer, who is employed and paid by the employer, can truly act "neutrally" when making determinations, as is required under the 2017 editions. Contractors highlighted their concerns that surely it is human nature to act with an element of bias toward the hand that feeds you, even if unintentional or unconscious.

Take the example of engineers who have produced contract designs at an earlier stage and are later required (in their subsequent position as the Engineer under the contract) to determine whether a variation should be instructed because of defects in their own design. Does human nature cloud judgment and create an inherent tension in this situation? On the other hand, others expressed the view that, in practice, an engineer who has been involved in the design process is often far better positioned to resolve issues relating to poor design between the contractor and the employer, as their knowledge can be invaluable.

The 2017 editions of the FIDIC works contracts recognise this tension and explicitly require the Engineer to act "neutrally" - new terminology compared to the 1999 first editions. In addition, the 2017 forms introduce a prohibition on any requirement for the Engineer to obtain the employer's consent before making a determination under sub-clause 3.7. While the contract can only perform as well as the people that use it (and maybe also those that amend it), the obligations on the Engineer, combined with a multi-tiered dispute resolution mechanism (and a dispute avoidance and adjudication board (DAAB)) should provide the checks and balances required. Certainly FIDIC show no signs of moving away from this approach in the RB and YB contracts.

The best way to address these issues is training. Another recurring theme at the conference was a call for contractors, employers and engineers to be properly trained on how to use and operate FIDIC contracts. If the drafting requires impartiality or neutrality, typically it is human failing in operating the contract that leads to problems. Knowledge of the contracts is paramount – not just for the engineer, but for employers and contractors too.

However, if parties are not able to understand how the contract works, an alternative is to simplify them. We have already seen from past conferences (and as was oft repeated at this year's conference) how FIDIC's short form Green Book contract, 2021 edition, has been positively received by the market. Questions were raised by delegates as to whether any future revision of the RB, YB and silver book (SB) contracts would take inspiration from the 2021 Green Book and perhaps move away from the very sophisticated and comprehensive, but admittedly fairly complicated, 2017 editions. It's an interesting proposition, particularly given the market's persisting attachment to the 1999 forms.

For more information about the 2021 Green Book, see Legal update, Blog post: FIDIC contracts – a preview of what is to come.

For more information about the role of the Engineer, see Practice note, FIDIC Red Book (2017): The Engineer.

New contract forms

The big teaser from the Contracts Committee at the conference is that they are considering changes to the RB, YB and SB contracts with "Objective 2030" revealed. Whether this will be a whole new edition or an update to the 2017 editions remains to be seen. Heads of Terms are to be established in 2026 and a Task Group to follow. FIDIC will be looking for feedback from the industry – so watch this space for surveys.

In the much nearer term, FIDIC anticipates that the following will be released in 2026:

• Golden Principles for the Services Contracts (White Book).
• Bronze Book test edition
• Several sub-contracts including for SB 1999 (evidencing the continued popularity of the 2017 suite's predecessor) and a draft sub-contract for the 2021 Green Book.
• Collaborative contract.

I wrote in detail about this contract following last year's conference so I won't repeat myself, except to note that the collaborative contract will still be recognisable as a FIDIC contract. It is not intended to be "soft and aspirational" but "smart and enforceable" with clear contractual obligations (and defaults). It is worth mentioning that through the "collaborative management team", the parties, while not relieving the Engineer of its duties, may be able to alleviate some of the reservations previously highlighted regarding this role. It looks like it will be a real coming together of people in the management of projects, which is exactly what many FIDIC contract users are asking for.

The EPCM contract is further down the line. There was no mention of the amendments for off-shore wind contracts (previously expected as a "plug-in" to the YB 2017).

This year's conference saw the release of a new Practice Note III: Dispute Board Decisions: Preparation and Composition, which provides guidance to dispute board members in reaching and writing their decisions.

For more information, see Article, FIDIC contracts – conference, carbon clauses and collaborative contracting and Legal update, Dispute boards: FIDIC practice note on dispute board decisions.

People making the difference – Carbon Management Guide

This year's conference also saw the release of FIDIC's Carbon Management Guide (CM Guide), which sets out FIDIC's overarching principles and approach to carbon management throughout the lifecycle of a project. The intention is that the CM Guide will be complemented by a separate Carbon Emissions Management (CEM) Guidance and annex for each FIDIC contract, which will set out contract-specific mechanisms and amendments. Conference delegates were given a pre-release of the CM Guide and the CEM Guidance to the White Book Services Contract. Both documents together inform each contract's approach to carbon management, with a focus on regular monitoring, reporting, damages and incentives.

The CM Guide however is not prescriptive, but gives examples and guidelines with suggested definitions and optional provisions. The aim is to make carbon a measured item, linked to performance. It is a practical tool allowing contract users to take action and measure tangible results, rather than have holistic aims without contractual measurement.

Eventually it is intended that adopters of the CM Guide will allocate performance damages if contractors fail to achieve the "Carbon Budget" (always within the overall limit of liability). However, the Contracts Committee are keen to avoid discouraging users on this basis. Performance damages are optional. Indeed, the message was very much that the CM Guide is flexible enough to implement in any project stage and that parties should not be afraid to take the initial step, even incorporating it part way through projects (without affecting existing risk profiles). FIDIC is urging the industry to start including basic principles in their contracts as soon as possible, even without incentives or damages, in order to begin to measure, to discuss and to monitor. As only with familiarity will adoption become commonplace.

Feedback from the multi-lateral development banks (MDBs) present is that they will be looking to incorporate carbon management principles into their bidding documents alongside existing environmental and social requirements. The focus is on qualitative criteria for contractors, with some MDBs already raising qualitative aspects of contract award to 50% minimum of tender criteria. Contractor delegates were also broadly positive, highlighting the competitive advantages to be gained from those trained and able to demonstrate familiarity with such practices and future proofing their businesses. People it seems are willing to make the jump, once they have the tools to enable them to do so. Let's hope the CM Guide can lead the way.

For more information, see Legal update, FIDIC launches Carbon Management Guide.

For more information generally about climate change provisions, see Practice note, Climate change clauses for construction contracts.

AI – the future and the present

Even at a conference with a clear spotlight on the human impact on project delivery, one of the most compelling sessions looked at the impact of AI on FIDIC contracts, in particular its use in contract management, project delivery and dispute resolution and balancing AI outputs with human judgment.

One example is the partnership between the American Arbitration Association - International Centre for Dispute Resolution (AAA-ICDR) and QuantumBlack AI by McKinsey to develop and deploy an "AI arbitrator" aimed at transforming arbitration in construction disputes, which was unveiled in November 2025. Having been fed 1500 construction arbitration awards, the AI arbitrator was invited to give decisions in bi-party, low value disputes, reportedly with some success. Crucially, these decisions are still overseen by a human arbitrator, yet the process is reported to achieve a 30% cost saving and a 25% faster process. It gives a taster of what might be expected from AI in construction disputes in the near future.

For FIDIC contract users this presents an interesting series of conundrums, starting with the type of decisions that should be fed into the model. FIDIC is an international contract and clearly jurisdiction-relevant decisions should be included, but what about decisions with similar facts based on the same FIDIC forms in different jurisdictions? Even if they are not binding precedent, a human arbitrator would no doubt be aware of such decisions and how FIDIC clauses are interpreted in different jurisdictions. Can AI now (or be able to in the future) consider different weighting of precedent decisions?

Another important question is whether FIDIC should adopt AI into its multi-tiered dispute resolution mechanism and if so at what level? Should it be limited to the arbitral process or be utilised for every determination under the contract; by the Engineer, the DAAB and arbitrators? The majority human response at the conference was "not at any level, yet", closely followed by "yes, at every stage of the mechanism". It would seem that humans are all-in on this, but only once AI capabilities have progressed sufficiently. When this is likely to be is a much bigger question.

Of course, there were other areas where AI is already being used very effectively and many more of us will have already experienced some of these opportunities with document review, summations and translations (though I hastily add not in the drafting of this article!).

For more information on the in AAA-ICDR AI arbitrator, see Legal update, AAA-ICDR will launch AI arbitrator for construction arbitrations in 2025.

For more information about AI generally, see AI toolkit (UK).

Conclusions and looking forward

The keen interest in the CM Guide and collaborative contract show that people want to co-operate with each other to deliver successful projects. Knowledge of the contracts was identified as the key to success with the parties' attitude and approach more important than contract clauses.

It is a shame that the collaborative contract is still not available for release, but its delay has in fact given pause for reflection. This conference demonstrated that there is an appetite for it among FIDIC contract users globally, many of whom value human engagement and contract management over the contract terms themselves.

We may have started with the premise that projects aren't delivered by the contracts we write, but by the people that bring them to life. However FIDIC consistently recognise that in order for people to behave properly, the contract must be fair and balanced and apportion risk to the party best able to manage it. This is a fundamental requirement of construction law: it is in no-one's interest to do otherwise as this risks insolvencies, mis-management and a claims culture. Whether the underlying contract is short form, long form, collaborative or otherwise, it is only by choosing the right form of contract with an appropriate risk allocation (which the parties understand and are familiar with) that we enable effective contract management and constructive human behaviours to deliver successful projects.

A version of this article was published in PLC Construction on 5 January 2026