The construction industry has always been built around managing uncertainty. 

Projects involve complex relationships, competing priorities, commercial pressures, and decisions that carry significant financial consequences. 

Artificial intelligence is now introducing a new category of risk, one that does not always fit neatly within traditional contractual boundaries. 

AI is moving rapidly from the margins into everyday delivery. 

From automated design tools and predictive analytics to project management platforms, technology is changing how developments are designed, constructed, and maintained. 

For an industry where time, cost, safety, and commercial efficiency are critical, the potential benefits are significant. 

However, innovation also creates exposure. 

While technology advances quickly, the legal and contractual questions surrounding AI are only beginning to emerge. 

When AI Makes a Mistake, Who Carries the Risk?

The most critical issue is accountability. 

If an AI system produces a flawed design recommendation, fails to identify a site risk, or contributes to a programme delay, who is responsible? 

Is it the software provider?

The contractor who relied upon the output?

The professional consultant who incorporated it?

Or the employer who specified its use? 

These are not technical questions; they are contractual issues. 

Standard contract forms, including JCT and NEC, provide established frameworks for allocating responsibility. However, they were not drafted with widespread AI-assisted decision-making in mind.

The use of technology does not remove professional duties. 

Consultants and contractors remain bound to exercise appropriate skill, care, and judgment. The challenge is ensuring contracts clearly address how liability is distributed when technology becomes central to decision-making. 

AI Supports Professional Judgment, It Does Not Replace It

There is a growing temptation to view AI as a substitute for professional expertise. 

It is not.

AI can process vast volumes of data and identify patterns at speed, but its outputs still require human verification, professional judgement and contractual oversight.

Crucially, it cannot take contractual responsibility when an issue arises. 

The future belongs to organisations that combine technological capability with human expertise and effective risk management. 

Data Is Becoming a Contractual Risk

AI systems depend entirely on data. 

Modern developments generate vast amounts of valuable data, including site surveys, BIM models, specifications, programmes and cost information.

As this information is shared with third-party platforms, crucial questions arise: 

• Who controls, uses, and reuses project data? 

• How is confidential commercial information protected? 

• Who owns AI-generated outputs created using project data? 

Without clear contractual provisions, data uncertainty will become a key foundation for future disputes. 

Closing the Contractual Gaps

Many construction agreements currently in use were developed before generative AI became a operational tool. As adoption increases, businesses must ensure their contracts properly address: 

• Responsibility for automated designs and analysis 

• Verification and sign-off processes for AI outputs 

• Liability associated with software or platform failures 
• Data ownership, confidentiality, and permitted use 

The organisations that benefit most from AI will be those that embrace innovation while maintaining clear governance, robust verification processes and well-drafted contracts.

AI has enormous potential to improve how projects are delivered, but every technological advancement changes how risk should be allocated.

As AI becomes more embedded across design, procurement and project delivery, should standard construction contracts evolve to deal with AI explicitly, or are the existing principles of contractual responsibility sufficient?

I would be interested to hear how others in the industry are approaching this.