In your recent piece for The International Construction Law Review, you explored how technology is transforming the delivery of major infrastructure projects—and the legal and practical challenges that come with it. Where are you seeing the most significant changes in the construction and infrastructure space today?
Honestly, it’s everywhere—but particularly in the construction industry, which is the space I know best. Construction has always pushed the envelope—an example is what we call “value engineering”— innovating to optimise outcomes by keeping costs low while enhancing project value and exploring alternative designs, materials etc. Today, I see the most significant shifts in three areas: renewables, the digital economy, and grid infrastructure. For example, in developed countries, our grids are outdated and need massive investment in expansion and upgrades. They were designed to radiate from traditional thermal power sources like coalfields or ports for oil and gas imports —not for today’s energy mix of renewables, small nuclear reactors, hydrogen, or ammonia. That shift requires us to redesign networks entirely.
You’ve used the phrase “rapid innovation” before. What does that mean in this context?
We’re in a moment where demand is outpacing supply. Everyone wants the latest tech—now. That’s forcing companies to push products and systems to market before they’re fully tested. Instead of derisking the application of new technologies through extended testing in controlled factory environments, we are learning how they perform live once they are installed on site. We then find they need to be fixed in challenging environments —like a construction site in the North Sea – far from the resources needed to perform the fix and invariably in jurisdictions with more stringent controls over labor conditions. This race to deploy faster comes at a price.
That kind of speed must have implications for construction standards. How do you define “state of the art” when technology moves so fast?
That’s exactly the issue. Contracts often use terms like “state of the art” or “world-class,” but what do those mean when today’s best is outdated by tomorrow? A great example is a UK dispute where offshore wind turbines were designed using proven design standards developed for oil platforms. But turbines are dynamic structures – the blades rotate; platforms don’t. A flaw in the design by a factor of ten meant that foundations that were supposed to last 20 years, failed in under two. The lesson? Whether it’s new tech or existing tech used in new ways, the margins for error are small and the potential exposures are huge.
Let’s talk about the contracts. How is this wave of tech disruption changing legal agreements?
They need to evolve, but it requires the industry to move out of the comfort zone of the known risk allocations of existing procurement models and standard form contracts so many haven’t caught up with the practical realities experienced on these projects. The legal system moves slowly, while technology moves fast. For example, the records relied on in resolving disputes used to be formal project correspondence, reports or meeting notes. Now, so much of the day to day communication about what is happening on site is through WhatsApp messages or common records stored on shared databases. Offshore wind is another example: early contracts adapted contracts for onshore wind projects with a few clauses thrown in to deal with the marine spread (the vessels). But offshore is completely different—the project is driven by vessels, weather and local environmental issues—all of it changes how risk needs to be managed. Contracts need to reflect that reality.
What about in the digital economy—like data centres?
That’s where we really need to think differently. If a data centre fails and the end user cannot access its data, it is catastrophic. We have become so reliant on instant access to data on computers and have seen how power failures shut down Heathrow airport and the Iberian peninsula’s economy. When that happens, its no longer a fight about who bears the cost of fixing what’s broken; its about who bears the consequential losses like lost profits etc. – the very costs that most standard form contracts seek to exclude. We need to move away from focusing on a “repair obligation” mindset and focus on fail-safes and redundancy. Before you hand over your data, you want near-zero failure probability. That demands a very different contractual regime and performance standard.
How do you advise clients when these risks materialise and you’re already in a dispute?
By the time I’m involved, it’s about looking at the contract you have signed up to in order to see what tools might be available. But even then, it is important to remember that the best legal answer isn’t always the best commercial one. Take liquidated damages: once a contractor hits the cap, they’re no longer incentivised to fix the problem quickly; they’re incentivised to do the minimum necessary to avoid being in wilful default (which would blow the protection of their caps on liability) and keep their costs down. Owners want the asset operational ASAP—that’s when the money flows. So we look for ways to realign incentives—even within a dispute.
There’s an irony here: we’re using technology to move faster, but people seem more risk-averse than ever.
Exactly. The pace of innovation is outstripping our ability to pause and assess.
An interesting example is the global conversation about who can fast-track and simplify connections to the grid access – whether it is for renewable energy projects or to become hubs for data centres and AI infrastructure? In the UK, for example, if the regulatory reforms announced earlier this year could by pass the queues blocked with “zombie projects” and so reduce a 10–15 year wait for grid connections, we’d be better positioned to lead in this space. It’s a window of opportunity—but one that won’t stay open long.
We can’t talk about this without talking about AI. What role do you see it playing in disputes?
AI is already helping us sift through massive datasets, find precedents, and identify key documents in huge trial bundles. What it won’t do—at least not yet—is replace judgment. Our cases are often first of a kind so the issues are unique, and the solutions need to be bespoke. We can draw on our back catalogue of experience but there is no pre-existing playbook that will solve all issues. But could we see disputes arising from AI-based decisions in construction? Absolutely. And already, with drone data and modeling tech, the definition of “good industry practice” is evolving—just like US constitutional interpretation, the words do not change but the understanding of what they might encompass is different.
What about the impact of all this on international and cross-border disputes?
COVID changed everything. Before, it was unthinkable to hold hearings remotely. Now, procedural hearings are almost always virtual. It saves time and cost, though it’s not without challenges—time zones, connectivity, and so on. That said, evidentiary hearings and mediations have largely returned to in-person formats, because you need to look your counterparty in the eye and decide whether to trust them or see the tell signs to work out whether they are telling the truth.
To close—what advice would you give to developers and project sponsors navigating these trends?
Ask yourself: do you want the newest technology or the most proven one? You can rarely have both; it’s a trade off. Set realistic performance criteria and understand your power needs—grid access, reliability, and the energy mix all matter. Bottom line, you get what you pay for. If you choose a faster or cheaper option, you need to accept that it may require more rework and oversight requirements. Whereas if you pay for premium engineering, it’ll probably take longer and cost more—but your chance of it working first time are better.
