Construction disputes are characterized by complexity, volume and high financial stakes. AI-enabled tools are now being deployed to help dispute boards ("DBs") manage that complexity and volume more efficiently. This alert analyzes the Dispute Resolution Board Foundation ("DRBF") Guidelines on the use of AI, which reflect a clear industry focus on efficiency and the need for robust governance frameworks to ensure responsible use of AI.
Artificial intelligence: Reshaping construction dispute administration
The digitization of the DB’s functions commenced well before the advent of AI. Remote dispute resolution, virtual inspections, drone-based or 3D site inspections and Building Information Modelling ("BIM")1 are now established features of complex construction projects. The 2022 Reprints of the FIDIC Red, Yellow and Silver Books have formalized this shift, expressly permitting DB meetings and site visits to be held online in appropriate circumstances.2
AI now stands at the forefront of this trend. In 2024, the DRBF published its Best Practice Guidelines for the use of Artificial Intelligence by Dispute Boards ("DRBF Guidelines") that guide DBs and users on how AI tools are likely to be used and how they should be managed responsibly to maximize efficiency and productivity, particularly when handling the complex and voluminous documentation that characterizes construction disputes.3
What the DRBF guidelines permit and prohibit
The DRBF Guidelines identify the following as the primary efficiency-enhancing applications of AI for stakeholders in DB proceedings:
- Reviewing and summarizing complex and voluminous documents and submissions;
- Conducting legal research;
- Reviewing delay and quantum claims submitted by the parties;
- Creating meeting agendas from correspondence and submissions;
- Performing data analysis; and
- Facilitating document translation.
The DRBF Guidelines govern the use of AI in DB proceedings through four key principles:
- Accountability: DBs are responsible for their processes and decisions, irrespective of whether or not AI was used, and cannot delegate any part of the decision-making to generative AI;
- Confidentiality and security: DBs must protect the confidentiality of the project, the parties, and site data at all times, notably with regard to the use of public or open AI platforms;
- Ethics: AI outputs may contain bias, errors, or hallucinations, requiring DBs to review AI outputs thoroughly and not treat them as substitutes for judgment, contextual understanding, and common-sense reasoning;
- Transparency: All stakeholders in DB proceedings must disclose any intended AI-use, and DBs must ensure a full understanding of how the parties have used AI tools, particularly for complex delay and quantum issues.
AI tools can assist DB members in accessing and processing real-time information about project conditions, costs, and timelines, which is a significant advantage in long-term, data-intensive projects.4 The 2025 International Arbitration Survey confirms this direction of progress: a majority of respondents already use AI tools for research, data analysis, and document review, and at least 90% expect to do so within five years, with time savings cited as the primary driver.5
The DRBF is unambiguous that AI must not be used to make decisions.6 A DB member is always responsible for their decision regardless of whether AI tools were used in its preparation. AI tools can be used for support, but they cannot substitute a DB member’s judgment, contextual understanding, or professional expertise.
In practice: Smart dispute boards in complex infrastructure projects
In practice, the use of AI tools as contemplated by the DRBF Guidelines has greatest potential in large-scale complex infrastructure projects such as shipping yards, oil and gas field developments, transport and urban development schemes, energy transition projects, including hydropower projects and more recently AI data center builds.
AI data centers require larger footprints and higher rack densities than traditional facilities. This can generate supply chain pressures that can rapidly escalate into significant disputes. The demand for AI data centers worldwide is on the rise, with market value projections going from USD 147.3 billion in 2025 to USD 810.6 billion by 2033.7 Global hotspots for new data center construction include markets in Asia-Pacific, the Middle East, Africa, and Latin America, several of which favor the use of FIDIC-based contracts and the potential for the emergence of smart DBs would be high based on practitioners’ choices in their projects.8
Using AI tools through a human-AI partnership is key to unlocking their full potential. AI tools can be used to handle repetitive, data-intensive tasks, leaving judgment on nuanced legal and factual issues to the human decider.9 Real-time AI-assisted document review, voluminous data analysis, and AI-assisted meeting preparation, among other "smart" uses, can help interface risks between prefabrication, logistics, and installation contractors in large infrastructure projects.
DBs and party representatives that will leverage advanced analytics would potentially identify fault patterns and assess precedents faster, thereby delivering more consistent outcomes. Parties that can harness data effectively, stand to gain a significant advantage in a dispute resolution process.10
Yet despite the advantages of AI capabilities, key challenges also need to be factored in. Data privacy, security vulnerabilities, algorithmic bias, and lack of transparency (the "black box"11 problem identified in the DRBF Guidelines). Industry and technological publications also caution against "automation bias"12 flagging the possibility of embedding historical bias through training data, due to tendencies to accept AI outputs without adequate scrutiny. DBs and practitioners must ensure thorough verification of results along with compliance with applicable data protection laws, professional conduct standards and contractual frameworks.
The EU AI Act’s regulatory dimension
The EU AI Act, a comprehensive legislative framework for regulating AI in the EU, provides for a risk-based approach to classify AI systems into four tiers, i.e., unacceptable, high, limited, and minimal risk.13 Annex III classifies "AI systems intended to be used […] in researching and interpreting facts and the law and in applying the law to a concrete set of facts, or to be used in a similar way in alternative dispute resolution" as high-risk.14 Recital 61 adds that high-risk should also include "outcomes of the alternative dispute resolution proceedings [that] produce legal effects for the parties".15
It remains to be seen whether smart DBs may fall within the ambit of these provisions. AI tools that support extracting and clustering relevant facts, selecting precedents and/or suggesting how to apply them to the facts could be in the high-risk zone. By contrast, AI systems that perform speech-to-text conversion, manage and search large data-sets, produce chronologies and/or carry out ancillary tasks such as document anonymization or language editing could fall outside this high-risk purview.
DB members, as natural persons using AI in their professional engagement, are likely "deployers" within Article 3(4) of the Act and therefore subject to its obligations when deploying high-risk systems.16 As explained in our EU AI Act Handbook, the Act applies extraterritorially to any deployer established or located in the EU, and to any deployer in a third country where the output produced by the AI system is intended to be used in the EU – a potentially significant reach for internationally seated DBs.17
Compliance obligations under the EU AI Act for stand-alone high-risk AI systems have been deferred and now apply only from 2 December 2027.18 This provides DB members and institutions with a window to prepare. As also noted in our AI Watch: Global regulatory tracker, the AI regulatory landscape is evolving rapidly and smart DBs dealing with international projects should monitor all applicable rules regulating the use of AI across jurisdictions.
Conclusion
AI tools can offer genuine and significant efficiency gains in DB proceedings and the DRBF Guidelines shine a light on how governance principles and responsible implementation are necessary. Human judgment and complete oversight are non-negotiable.
For construction companies delivering complex, high-value projects, the integration of AI into DB proceedings is an operational reality that calls for proactive engagement. This includes better structuring DB provisions at the contract drafting stage, selecting AI-ready governance frameworks, and effectively managing large datasets with AI tools in live disputes. Stakeholders investing in deploying smart DB tools now will be better positioned to manage disputes more efficiently, within regulatory bounds, at a lower cost, and with enforceable outcomes in the long run.
Natalia Gracia Gómez (Stagiaire) and Maher Abdelaziz (Jurist) contributed to the development of this publication.
1 BIM is a digital modelling process that integrates design, construction, and project management data across all stakeholders, supporting collaboration and process automation.
2 White & Case, "Important Changes to FIDIC’s Rainbow Suite", 24 February 2023, available here.
3 DRBF, "Best Practice Guidelines for the use of Artificial Intelligence by Dispute Boards", 7 November 2024, available here.
4 A recent study found that hybrid decisions generated using AI achieved a 92.4% agreement with human outcomes. The study also notes that average case resolution time was reduced from 48 hours to approximately 6.5 minutes (human-AI hybrid), a reduction of 99.5% compared to manual arbitration processes. See Ping Han, "AI-powered digital arbitration framework leveraging smart contracts and electronic evidence authentication", Scientific Reports, 24 October 2025, available here. The authors note that this is a single study within a larger research trajectory and success rates will likely not be as high with real-world construction arbitration datasets.
5 David Robertson, Kevin Touhey, "Construction arbitration could be speeding up – here is why", White & Case, 18 December 2025, available here, citing the 2025 International Arbitration Survey conducted by the School of International Arbitration at Queen Mary University of London and White & Case, available here. The Survey compiles responses from more than 2,400 practitioners, arbitrators, arbitral institutions, academics and clients.
6 DRBF, Revised Code of Ethical Conduct, October 2024, available here. Canon 4 provides that "Dispute Board recommendations, determinations and decisions must be made expeditiously and personally by the members of the Board itself based on the provisions of the contract, the applicable law and the information, facts and circumstances submitted by the contracting parties" (emphasis added).
7 Next Generation Technologies Research Team, "AI Data Center Market (2026 - 2033)", Grand View Research, June 2026, available here; PR Newswire, "AI Data Center Market Projected to Reach USD 810.6 Billion by 2033 as Enterprises Accelerate Investments in AI Infrastructure", 23 June 2026, available here. See also Adam Cieply, Jennifer Iacono, James Johnson, Eric Klar, "Buy the power: Data center deals on the rise in the US", White & Case M&A Explorer, 7 March 2025, available here.
8 Brody K. Greenwald, Darryl Lew, Kate Perumal, Efat Elsherif, "Server wars: the future of data center arbitration", White & Case, 20 August 2025, available here; Paddy Mohen, Fola Oginni, "Intelligent Design: Constructing next generation data centers for the AI boom", White & Case, 19 February 2025, available here.
9 Effective integration requires clear protocols, ongoing training, and a commitment to transparency and accountability. See Jean-Rémi de Maistre, Tiffany Lam, "The Human-AI Partnership in Arbitration: Lessons from China Arbitration Week", Daily Jus, 9 October 2025, available here; Annie Lespérance, "Data Wars: How Legal Data & Specialized AI are the New Competitive Advantage in Arbitration", Daily Jus, 22 October 2025, available here.
10 White & Case, "Statistical Sampling in Construction Disputes: balancing and managing the risk", 28 March 2023, available here. Result-predictive tools, such as the AAA Resolution Simulator, can also help parties in refining their legal or negotiation strategy, set realistic expectations, and/or evaluate whether alternative resolution paths might be appropriate. See Bob Ambrogi, "American Arbitration Association Launches Resolution Simulator, Expanding Its AI Arbitrator Tool", LawSites, 4 March 2026, available here.
11 This refers to the difficulty of understanding why an AI system reached a particular output. See DRBF, Best Practice Guidelines for the Use of Artificial Intelligence, November 2024, Glossary: "Black Box Syndrome," available here.
12 This refers to the tendency to over-rely on automated outputs without sufficient critical scrutiny.
13 See further European Commission, "Draft Commission guidelines on the classification of high-risk AI systems", 19 May 2026, available here.
14 Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (EU AI Act), 12 July 2024, Article 6(2), Annex III, 8(a) ("AI systems intended to be used by a judicial authority or on their behalf to assist a judicial authority in researching and interpreting facts and the law and in applying the law to a concrete set of facts, or to be used in a similar way in alternative dispute resolution"), available here. Notably, high-risk systems are subject to the most demanding obligations, including requirements on data governance, technical documentation, human oversight, and cybersecurity.
15 See Maxi Scherer, "We Still Need to Talk About the EU AI Act - and Before 23 July Now the Draft High-Risk Guidelines Are Here", 2 July 2026, Kluwer Arbitration Blog, available here; Maxi Scherer, Veronika Pavlovskaya, "Why the LCIA and Other Leading Arbitral Institutions Are Speaking Up on the EU AI Act", 16 July 2026, available here.
16 Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (EU AI Act), 12 July 2024, Article 3(4), Recital 13 ("The notion of ‘deployer’ referred to in this Regulation should be interpreted as any natural or legal person, including a public authority, agency or other body, using an AI system under its authority, except where the AI system is used in the course of a personal non-professional activity. Depending on the type of AI system, the use of the system may affect persons other than the deployer."), available here.
17 Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (EU AI Act), 12 July 2024, Recital 22 ("To prevent the circumvention of this Regulation and to ensure an effective protection of natural persons located in the Union, this Regulation should also apply to providers and deployers of AI systems that are established in a third country, to the extent the output produced by those systems is intended to be used in the Union."), available here.
18 EU Council, Press Release, "Artificial Intelligence: Council and Parliament agree to simplify and streamline rules", 7 May 2026, available here.
White & Case means the international legal practice comprising White & Case LLP, a New York State registered limited liability partnership, White & Case LLP, a limited liability partnership incorporated under English law and all other affiliated partnerships, companies and entities.
This article is prepared for the general information of interested persons. It is not, and does not attempt to be, comprehensive in nature. Due to the general nature of its content, it should not be regarded as legal advice.
© 2026 White & Case LLP
