Artificial Intelligence and Construction Contracts

Artificial Intelligence (AI) is a rapidly evolving technology that is being used by the construction industry at different points during the project lifecycle. For example, AI is being used for planning, design, estimating, bidding, negotiating, drafting construction documents, project management, scheduling, productivity, resource allocation, risk management, site safety, regulatory compliance, communications, quality control, robotics, predictive equipment maintenance, and dispute resolution. The use of AI, however, is not risk-free, is still in its infancy in many ways, and presents unique challenges.

This article provides a definition of AI, discusses some of the risks associated with using AI, and identifies best practices for addressing AI in construction contracts.

What is AI?

There are many definitions of AI. For example, Google defines AI as “a set of technologies that enable computers to perform a variety of advanced functions, including the ability to see, understand and translate spoken and written language, analyze data, make recommendations, and more.”[1] IBM defines AI as “technology that enables computers and machines to simulate human learning, comprehension, problem solving, decision making, creativity and autonomy.”[2]

When discussing AI, Industry experts also draw a distinction between traditional or “non-generative” AI and “generative” AI. Traditional or Non-generative AI “concentrates on categorizing and evaluating preexisting data rather than producing original content. It performs effectively in tasks requiring pattern recognition and pattern-based prediction.”[3] Generative AI, on the other hand, is defined as AI “that can create original content such as text, images, video, audio or software code in response to a user’s prompt or request.”[4] Generative AI “relies on sophisticated machine learning models called deep learning models, algorithms that simulate the learning and decision-making of the human brain.”[5]

It is also important to understand the differences between “open-source” AI and “closed-source” AI. The Open Source Initiative defines Open-Source AIAS an AI system made available under terms and in a way that grants the freedoms to: (1) use the system for any purpose and without having to ask for permission; (2) study how the system works and inspect its components; (3) modify the system for any purpose, including to change its output; and (4) share the system for others to use with or without modifications, for any purpose.[6] Closed-Source AI is proprietary and “refers to artificial intelligence technology where the source code, algorithms, and inner workings are kept private and not made publicly available.”[7]

There are risks and ethical concerns when using AI. Think of it as “garbage in, garbage out.” AI depends on the accuracy of the databases or datasets it uses. If those datasets are biased, discriminatory, incomplete, or erroneous, the information or answers produced by AI may also be biased, discriminatory, incomplete, or erroneous. AI may also “hallucinate” by generating outputs that are inaccurate, nonexistent, or nonsensical. The use of AI is also dependent on accurate data prompting. A human user must input queries and parameters into an AI system to yield usable, reliable results. Thus, human error or lack of training may also skew an AI output or result. In addition, AI is vulnerable to cybersecurity attacks and the technology itself can be used to conduct more efficient and effective cyberattacks. There are also privacy concerns with using AI and particularly open AI, as the information input into an open AI system is publicly available. In addition, the use of generative AI presents novel intellectual property issues. Ethical concerns include the elimination of jobs, increased needs for scarce resources such as water and electricity used to run AI systems, and elimination of human thought and creativity.

Given the risks inherent in the use of AI at this point in its evolution, it is critical that construction professionals and project owners be transparent about their intended use of AI on each construction project and add provisions to their construction contracts that address AI in a variety of contexts. These concepts are discussed in additional detail below.

Consider adding an AI transparency and disclosure clause to construction contracts that requires each party to be transparent about their use of AI and to disclose what AI products, tools, data sources, algorithms, and decision-making processes will be used on the project.

Construction professionals should consider including an AI definition, scope, and purpose clause in their construction contracts. These types of clauses may include a definition of what constitutes AI for purposes of the contract and identify the agreed upon scope and purpose of the AI systems to be used on the project, the system’s functions and limitations, and parameters for human interaction with the system. A “human in the loop” is essential to AI use and management and may be required to meet the standard of care in performing the work.

Construction professionals will want to include AI liability and indemnification provisions in their construction contracts. These provisions should clearly allocate the risks that may arise from using AI between the parties.

Construction professionals should address AI data privacy and security risks and obligations in their construction contracts. These types of clauses should identify steps to be taken to minimize data privacy and security risks, ensure compliance with data protection laws, and delineate how the parties will deal with unauthorized access or data breaches.

Congress originally created the U.S. patent, trademark, and copyright systems to protect and encourage human creativity and invention. As a result, questions have arisen about how the federal government and federal courts will analyze intellectual property ownership as AI systems, including generative AI, play a greater role in the innovation process. Therefore, it is critical that construction professionals address  AI-related intellectual property ownership in their construction contracts. For example, the contract should clearly state whether intellectual property rights belong to the AI developer, the user, or a third party. The parties should also familiarize themselves with licensing and usage rights associated with the AI systems to be used on the project and address them in the contract, if needed.

The parties to a construction contract should also address AI regulatory compliance, carefully allocating the risk of compliance with AI laws, which may vary from state to state, between the parties. For example, in 2024, the Colorado legislature became the first state to enact laws governing Consumer Protections for Artificial Intelligence and established a regulatory framework aimed at lowering the risk of algorithmic discrimination in artificial intelligence-based decision-making technology.[8] The law requires certain businesses and employers using AI to make disclosures and meet other requirements. A growing number of states and the federal government have also passed and/or are working on similar legislation.

As part of a construction contract provision that addresses termination and post-termination rights, construction professionals should also address AI. For instance, the contract should address whether an AI system failure constitutes a breach of the contract, the parties’ obligations to maintain AI data and information post-termination, and AI system audits and review rights.

As part of a construction contract’s dispute resolution clause, the parties should also address the use of AI in the dispute resolution process. In determining whether to require arbitration or resolve disputes in court, parties should consider whether a third-party AI developer can be joined in arbitration proceedings in the event of an AI dispute. The parties may also wish to address whether AI may be used by the parties to present the dispute itself to an arbitrator, what will happen in the event one of the parties submits AI-hallucinated case law or other misinformation to the arbitrator, and whether the arbitrator may use AI to assist the arbitrator in deciding the dispute.  

BHGR’s Construction Group assists our clients efficiently and effectively throughout the lifespan of their construction projects. Applicable here, our attorneys negotiate, draft, and review all types of construction contracts, including EPC, design-build, design-bid-build, CMAR, and concession and development contracts, as well as drop-down and lower-tier subcontracts, framework, financing, joint venture, master services, secondment, and supply agreements.

[1] https://cloud.google.com/learn/what-is-artificial-intelligence.

[2] https://www.ibm.com/think/topics/artificial-intelligence.

[3] https://aitooltalks.com/non-generative-ai/#what-is-non-generative-ai.

[4] https://www.ibm.com/think/topics/generative-ai.

[5] Id.

[6] https://opensource.org/ai/open-source-ai-definition.

[7] https://tensorwave.com/glossary/closed-source-ai

[8] https://leg.colorado.gov/sites/default/files/2024a_205_signed.pdf. To learn more, read Colorado Legislature Further Delays Implementation of Consumer Protections for Artificial Intelligence Law to June 30, 2026: What Employers Need to Know by BHGR.