Artificial Intelligence (AI) in UK Construction Industry. Transforming Health and Safety or Creating New Risks?

From autonomous robots patrolling sites to AI powered cameras monitoring worker behaviour, the technology is rapidly moving from science fiction to the reality of UK construction sites.

But what does this mean for health and safety? Is AI genuinely transforming how we protect workers, or is it creating new risks that we do not yet fully understand?

And perhaps most importantly, how should the construction industry, with its proud but often cautious approach to new technology respond?

In this blog, I want to look at how AI is being used in UK construction, what the benefits and challenges are, and what the future might hold for an industry that is simultaneously embracing and yet wrestling with this transformative technology.

What Is AI Doing on Construction Sites?

Artificial intelligence is already making significant inroads into construction, with contractors leveraging AI-powered tools to monitor worksites in real time, identify potential hazards and enforce safety protocols. The hope is that by automating multiple complex processes, AI could help reduce fatalities and serious injuries in a sector that continues to suffer relatively high incident rates.

Computer Vision and Hazard Detection

One of the most immediate applications of AI in construction is computer vision and  the use of AI powered cameras to analyse live video feeds from job sites.

These systems can identify unsafe behaviours, such as failure to wear appropriate personal protective equipment, workers entering exclusion zones, or unsafe plant movements. When a hazard is detected, automated alerts notify safety managers, enabling immediate intervention.

Authors thought: Should we not be rolling out AI to document and champion positive observations and good practice also? Surly it can’t just be about policing the work area?

Wearable Technology and Biometrics

AI is also being used to support wearable devices that can monitor workers' movements and postures as they perform tasks, giving detailed feedback on how to avoid the causes of musculoskeletal disorders. This is particularly significant given that more than half a million workers in Great Britain suffered from new or long-term MSDs in 2024/25, with more than seven million working days lost. Construction was among the hardest-hit sectors.

Authors thought: A good example of AI making a real difference.

Wearable sensors are now able to contain three dimensional motion sensors driven by AI which can recognise patterns that may correlate with musculoskeletal risk.

AI has been trained using data from international ergonomic guidelines, but it can also learn and offer personalised coaching.

Smart helmets and vests equipped with biometric sensors are also being used to monitor worker fatigue, heart rate and environmental exposure, with AI analysing the data to detect signs of exhaustion or hazardous conditions and alerting supervisors when risk thresholds are crossed.

Autonomous Robotics

Perhaps the most visually striking application of AI in construction is the deployment of autonomous robots. In July 2026, McLaren Construction partnered with a robotics specialist to deploy autonomous robots across its UK construction sites.

The robots capture 360-degree site imagery, generate point cloud data, support progress verification, and carry out safety compliance patrols and quality assurance.

Adam Nicholson, McLaren's group preconstruction director, noted:

"The significance of this deployment for the construction industry is that we can move beyond machines that are remote controlled or pre-programmed for a limited range of tasks and routes. Instead, we now have autonomous robots navigating stairs, doors and other obstacles and constantly working with our human teams to support productivity, safety and quality".

In April 2026, Tilbury Douglas became the first UK contractor to deploy a humanoid robot on a live construction site. The robot, named Douglas, was trialled for ten weeks to autonomously capture 360-degree imagery, produce progress reports and carry out health and safety monitoring.

AI-Assisted Risk Assessment

Another emerging trend with significant implications for health and safety professionals is the use of large language models, such as ChatGPT, to generate risk assessments and method statements. Upload a picture of a situation on a job site, for example, a trench or scaffold and with the right prompts, the software can generate a fairly comprehensive risk assessment in a few seconds.

Author note: I reject more risk assessments now that I ever have in the past and that’s because contractors are using AI to write risk assessments that are neither fit for purpose or specific to the risks or project that they are being written for.

Some app’s allow workers to record 30-second video of their site, explaining hazards that are present or absent. The app’s AI engine reviews the data and generates a visual risk assessment, including a bullet-point list of identified risks.

In March 2026, Amey began rolling an AI enabled risk assessment tool across more than 2,500 field workers following a successful pilot.

The Benefits – What AI Is Getting Right

Predictive Risk Management

AI systems can analyse historical workplace data, including accident reports and environmental conditions, to predict potential safety hazards. This moves safety management from a reactive to a proactive approach, identifying risks before they materialise rather than responding after incidents occur.

Consistency and Speed

AI does not get tired, distracted, or influenced by familiarity. It applies the same standards consistently, every time. It can process vast amounts of data far more quickly than any human, identifying patterns and anomalies that might otherwise go unnoticed.

The HSE is itself using AI language models to understand risk profiles across the construction sector. Sarah Newton, Chair of the HSE Board, explained:

"This helps us to identify mismatches between the actual hazards causing workplace incidents and the risks our inspectors commonly focus on during site visits".

The Challenges.

The Trust Gap Between Management and Workers

Perhaps the most significant challenge facing AI adoption in construction is the gap between management enthusiasm and worker scepticism.

According to the Rapid Global Workplace Safety 2025 Report, 51% of managers said their business was ready to adopt AI to improve safety, compared to just 20% of frontline workers.

The majority of managers (75%) were confident AI could cut risk and improve safety, compared to less than half of workers at 44%.

This trust gap is not unfounded. Nearly one in five workers (18%) said safety processes caused them anxiety, and a further 60% said their employer did not support positive mental health.

If workers feel that AI is being used to monitor them rather than protect them, resistance will be significant.

Accuracy and Hallucination Risks

Large language models can produce plausible yet incorrect answers, a phenomenon known as "hallucination". When AI is used to generate risk assessments, the stakes are high. An incorrect or incomplete risk assessment could have fatal consequences.

Experts have warned that the construction industry has witnessed a surge in AI and mobile applications promising to streamline the creation of risk assessments and method statements.

While these technological solutions offer convenience, there are significant concerns about their accuracy and reliability.

Data Privacy and Ethics

AI has major implications for cyber security, privacy and ethics, particularly when it is used to monitor and analyse worker behaviour.

At the HSE's 2025 AGM, Rob McGill, national health and safety officer at Unite, raised concerns about what he described as excessive use of AI to monitor workers.

HSE chief executive Sarah Albon responded that monitoring workers was not "Big Brother" provided it was used for purposes of safety and that consultation before use had taken place.

Disconnected Data and System Fragmentation

Artificial intelligence is transforming workplace safety but disconnected data and poorly implemented systems risk creating more noise than insight. AI is only as good as the data it is trained on, and the construction industry has historically struggled with fragmented, inconsistent, and siloed data.

The Risk of Over-Reliance

There is a growing concern about over-reliance on technology. As AI becomes more embedded in the world of work, we must ask: will AI make work safer, or simply change the nature of the risks we face?

What the Industry Is Getting Right

Major Investment in AI Capability

Balfour Beatty's £7.2m investment in Microsoft 365 Copilot – one of the UK construction industry's largest known commitments to artificial intelligence signals that major players are taking AI seriously. The technology will support employees across hundreds of projects, with the aim of improving productivity and site safety through automation and better decision-making.

Collaborative Development

Amey has demonstrated how contractors and technology providers can work together to develop AI tools that meet the specific needs of frontline teams.

Regulatory Engagement

The HSE is actively engaging with AI, compiling a database of real-world examples across the industry sectors it regulates to aid the development of insights pertinent to the risk assessment of AI applications in industrial settings.

The regulator has also undertaken research into how businesses are adopting AI and its implications for workplace health and safety, examining the use of AI in developing health and safety provisions across a wide range of sectors, including construction.

Where There Is Still Room for Improvement

Training and Upskilling

Building engineers and construction professionals are being urged to upskill in artificial intelligence as the UK government warns that the construction and built environment sectors risk being left behind in the AI transition.

Training provision remains patchy, particularly for smaller firms.

Support for SMEs

For small and medium enterprises, AI is increasingly viewed not as an optional enhancement but as a strategic necessity for sustaining competitiveness, improving operational efficiency, and enhancing safety. However, these benefits are constrained by challenges including financial limitations, skills gaps, cultural resistance, technical barriers, strategic uncertainty, and security risks.

Incompatible Procurement Practices

The Institution of Civil Engineers has highlighted that incompatible procurement practices, data-hoarding, and weak governance are holding back AI uptake in the built environment sector.

Bridging the Worker-Manager Gap

The significant gap between management confidence in AI and worker scepticism needs to be addressed through visible proof that technology makes safety easier, training more consistent, and work sites safer.

What the Future Holds

A Single Construction Regulator

The journey towards a single construction regulator, which will bring together all the functions relating to the construction industry, is expected to continue. How this new regulator approaches AI, be it as an opportunity to be embraced or a risk to be managed will shape the industry's relationship with the technology for years to come.

AI and the Golden Thread

The Building Safety Act's requirement for a "golden thread" of information throughout a building's lifecycle is an area where AI can make a significant contribution.

Digital construction experts at Leeds Beckett University are already working to create a golden thread for fire prevention products, using AI predictive models and data monitoring to automate the identification of incorrect or faulty installations.

Continued Growth in Autonomous Systems

The use of autonomous robots on construction sites is likely to grow. Mace's Constructing with Intelligence report has called on UK Research and Innovation to establish a dedicated funding programme for robotics and artificial intelligence applications in construction, focusing on tasks where automation is most immediately viable.

Final Thoughts

Artificial intelligence is already transforming health and safety in the UK construction industry. From computer vision cameras that detect PPE failures to wearable sensors that prevent musculoskeletal injuries, from autonomous robots that patrol sites to AI generated risk assessments, the technology is here and it is not going away.

However. AI is not a magic bullet. It brings with it significant challenges including accuracy concerns, data privacy issues, the risk of over reliance, and perhaps most importantly, a trust gap between management and workers that must be bridged if the technology is to achieve its potential.

People must remain at the centre of everything. AI can highlight construction risks, but it still requires human intervention to resolve or reduce the risk.

The construction firms that will succeed in this new landscape will be those that treat AI as a tool to support their business, not to replace the expertise, judgement, and experience of their people.

The firms that do not and  treat AI as a quick fix or, worse, as a means of surveillance, will find themselves facing resistance, regulatory scrutiny, and ultimately, failure to deliver on the promise of safer, healthier construction sites.

If you require further assistance with any of the topics raised in this post whether that is understanding how AI might affect your health and safety management systems, developing a digital transformation strategy, or simply staying up to date with the latest developments.

Please get in touch: Yorkshire Health and Safety