Nick Burgess – Head of Professional & Management Risks at Markel UK considers AI’s impact on two significant insurance classes in the coming year.

Artificial intelligence is beginning to accelerate change in every sector of the economy. From precision forecasting to automation, AI promises major gains in speed and efficiency. But rapid progress brings risk, including in sectors like professional services and construction, areas of the economy that are particularly relevant to the Markel business.

AI risk is developing in the professional indemnity sector

Professionals using AI for client-facing work are facing new liability risks and we are likely to see greater signs of this in 2026. Accountants relying on automated tools for audits, architects using generative design software, management consultants relying on AI content are all now more exposed to errors from third party output.

When professionals rely on AI tools to generate work, they remain responsible for the outputs. If AI is used to produce tax advice that turns out to be inaccurate, and the professional fails to verify it, the liability sits squarely with them or their firm.

Professionals placing too much trust in AI-generated content, especially when outputs appear plausible and are delivered instantly, is a present and growing risk. Overconfidence can lead to missed red flags, not reviewing critical documents with sufficient diligence and increased potential for negligent advice.

A rise in litigant in person claims, where someone represents themselves in court without a solicitor or barrister, is another trend that could start to appear more strongly in 2026. This has the potential to increase claims volumes for insurers, while also exposing individual claimants taking this route to greater risk. A valid claim could fail if AI overlooks key legal nuances, something that might have been avoided with proper legal representation.

These scenarios will become more common and complex as AI develops and firms race to embrace the technology for fear of being left behind in a competitive landscape.

A complex risk landscape in the construction sector

In construction, AI-driven tools are now being used to optimise schedules, design buildings, assess structural risk and monitor on-site activity. With mass housebuilding and ever more complex commercial building structures, AI and other advanced software tools have arguably never been more welcome for the sector.

Whilst these tools improve productivity, however, they also introduce a new type of operational dependency, one that could have serious consequences if systems fail or data is inaccurate.

Errors driven by more reliance on AI could affect everything from project delays to unsafe building work. If decisions are made based on flawed data sets or incorrect software output, contractors could face new claims and insurers left grappling with liability claims that are difficult to resolve.

For larger construction projects, the complexity is compounded by the number of parties involved. Who owns the risk when AI output is relied upon or interpreted incorrectly by different parties? The contractor? The software provider? The subcontractor who took advice on face value? For underwriters, these blurred lines present a material challenge.

AI is changing the way people and organisations work and this will only accelerate in 2026. In the PI and construction sectors, AI will test how well insurers can manage an entirely new kind of risk that thinks for itself.