In our prior articles on artificial intelligence (AI) in construction, we discussed machine learning, image recognition, sensors-on-site, building information modeling, and smart contracts. As we noted, significant legal issues will arise with the increasing implementation of these technologies. These issues can be grouped generally into: (1) risk allocation; (2) ownership and protection of the technology, as well as the data input and outputs; and (3) the applicable standard of liability. In this concluding article of this series, we discuss those issues briefly below.

First, risk allocation. Specific contractual terms should be drafted to take into account which party is responsible for which aspects of the technology’s use. These terms should be negotiated fully between the parties, with knowledgeable counsel involved, and updated regularly. If a construction company has created its own technology, the risk of failure or mistakes will likely fall on the contractor. But if a contractor uses a third party’s technology, the parties will need to pay close attention to the apportionment of risk. The parties should address what duties each party owes to the others regarding the technology and its use. For example, the duty to observe and/or monitor the technology should be assigned. Related questions to this duty are how frequently, actively, and closely that monitoring needs to be performed. Finally, the parties should address the issue of a flaw in the technology being found. Is there a duty to disclose, who has it, and is there a companion duty to fix the issue?

Second, in general, ownership of any technology or algorithm rests with whoever created the technology. However, contractual arrangements can be put in place that allow for licensing and use of the technology. Such arrangements usually depend on whether a contractor developed the algorithm itself or if a third-party provider did. The same concerns emerge regarding the data – who owns it, who can access it, and how long must the owner hold on to it. An associated implication of the use of technology and data ownership is its protection and security. With greater use of connected devices and technology at worksites, contractors may open themselves to hacking by third-party bad actors. Therefore, privacy, confidentiality, and security are concerns that must be accounted for. Otherwise, determining responsibility for a breach could become very difficult and increase the likelihood of litigation.

Third, beyond determining who is responsible for injury arising from the use of AI, the companion question is the standard of liability that would apply. Would a strict liability regime — where the party responsible for the use of the AI would be liable without any need to prove negligence — or would a negligence standard be employed? An evolving trend among commentators is to suggest that strict liability will govern where fault cannot be determined. A major concern for contractors is whether the use of such technology will impose this higher level of responsibility.

Conclusion

Technological advancement and the implementation of AI in construction pose new legal implications and questions for industry stakeholders. While most concerns can be addressed through careful drafting of contracts, stakeholders should be aware of these new and different legal issues. Moreover, important elements of uncertainty exist regarding the legal standards, responsibilities, and expectations of parties when integrating this technology into construction. Nonetheless, early adopters potentially stand to gain a competitive advantage over others who lag behind. The lesson, as with construction generally, is careful contract drafting by counsel fully informed and knowledgeable of the cutting edge issues.