The Illusion of Imminence: Where is this Much-Fabled AI Revolution?

For much of the past decade, the UK property industry has spoken about Artificial Intelligence as something that is coming: imminent, inevitable, but perpetually just over the horizon. Conference agendas are crowded with PropTech panels, innovation strategies are filled with references to data and automation, and most large firms can point to at least a handful of pilots, lackluster initiatives or proofs of concept. And yet, for many surveyors on the ground, progress still feels tentative and fragmented. AI appears present everywhere in rhetoric, but only selectively embedded in reality.

This apparent apathy, however, masks a more nuanced truth. The issue is not that AI has failed to find relevance within property, nor that the technology itself is immature. In fact, AI is already delivering tangible value across transactions, valuation, asset management, and building operations. The real question is why adoption has been slower, more cautious, and more uneven than in many comparable industries. The answer lies less in algorithms and far more in the structural characteristics of the UK property sector itself.

To understand where the industry is heading, we must first be honest about where it currently stands.

The Shift: From Pilot to Workflow Integration

Over the past twelve to eighteen months, a meaningful shift has taken place. AI has moved decisively out of the experimental phase and into everyday professional workflows. Only a year ago, most initiatives were still framed as pilots, often confined to small innovation teams working at the periphery of the business. These efforts tended to focus on conversational tools; chatbots designed to answer basic questions, draft text, or surface information from document repositories. While useful, they were rarely connected to core operational processes.

Today, that picture looks markedly different. AI is increasingly embedded into the daily work of surveyors, analysts, asset managers, and operational teams. The speed of change has been striking. Whereas much of last year’s activity revolved around using Large Language Models (LLMs) as passive assistants, attention has now shifted towards agentic systems; AI agents capable of executing multi-step tasks autonomously, gathering information, validating inputs, drafting outputs, and escalating issues for human review at defined points.

Yet despite this acceleration, one principle has remained largely intact across the UK property sector: AI is being deployed as decision support, not as a decision maker. That distinction is not accidental, nor is it merely cultural conservatism. It reflects a deeply ingrained understanding of professional accountability and risk that continues to shape how far firms are willing, and able, to go.

AI in Surveying Practice: Augmentation, Not Automation

Transactions and Due Diligence

In practice, this becomes most evident in transactions and due diligence, which remain the most mature area of AI adoption. Here, the value proposition is clear and immediate. AI systems are now routinely used to read and analyse large document packs, extract key lease clauses, summarise planning conditions, EPCs, and operational manuals, and generate first drafts of leases, listings, and due diligence reports. The critical concept underpinning all of these use cases is that of the “first pass”. AI allows teams to surface issues faster, structure information more consistently, and focus professional attention where it matters most. It does not remove the need for professional judgement and experience; rather, it sharpens it.

Valuation and Market Research

A similar pattern can be observed in valuation and market research. AI is increasingly used to shortlist comparable evidence, draft initial market commentary, and run scenario or sensitivity analysis at a speed and scale that would previously have been impractical. However, the valuation opinion itself remains firmly (and legally) with the valuer. From both a professional and an insurance perspective, it cannot be otherwise. AI accelerates analysis, but it does not, and should not, sign off opinions of value. It never will.

Asset and Portfolio Management

In asset and portfolio management, the emphasis shifts again, from speed to perspective. AI enables firms to interrogate their portfolios in new ways, exploring questions around interest rate sensitivity, vacancy exposure, or capital allocation priorities with far greater depth and consistency than manual approaches allow. Once more, this is not automation of decision-making, but augmentation of strategic thinking.

Building Operations and Energy Management

Perhaps the clearest illustration of AI’s potential comes in building operations and energy management, where adoption has been strongest wherever good quality data exists. Predictive maintenance, energy optimisation, and early fault detection all lend themselves naturally to data-driven approaches, and the financial benefits are often straightforward to quantify. Unsurprisingly, these use cases have met less resistance than those touching core professional judgement.

The Structural Constraints on UK Adoption

Given this breadth of application, it is reasonable to ask why AI adoption in property does not feel further advanced. The answer is that the principal constraints are not technological. They are structural and human, as is often the case.

The Data Foundation Challenge

The most obvious, and most persistent, challenge is data. Property data is notoriously fragmented, inconsistent, expensive to access and often unstructured. The same asset may appear under multiple names(or addresses) across different systems; documents frequently contradict one another; critical information is often buried in siloed repositories of PDFs, scans, or long email chains. AI systems struggle to scale under these conditions. Without solid data foundations, even the most sophisticated models will underperform.

There is also a deeper, sector-specific issue at play. Property is fundamentally non-standardised. No two assets are truly alike. Physical characteristics vary, as do tenure structures, incentive packages, and contractual nuances. Unlike commodities or consumer goods, property transactions are high-value, low-volume, and inherently unique. This makes the creation of clean, statistically robust datasets far more difficult than in industries dealing with standardised products traded at scale.

While the UK has a rich property-related data landscape (and even richer contextual datasets), stronger than most EU counterparts, it has notorious gaps in datasets. There is no reliable leasing data, no API’s to HM Land Registry to extract mass sales transactions, fragmented planning data behind endless firewalls and local authorities poorly integrated at a national data level. While in some respects there is a risk of being overwhelmed with data, significant gaps remain which limits potential AI use cases.

Governance, Accountability, and Risk

Beyond data, questions of accountability, governance, and data protection loom large. Most of the activities within the property sector are tightly regulated and Governments are generally (often rightfully) slow to embrace change which might impact the public. When Artefact teaches practical AI courses through the Royal Institution of Chartered Surveyors (RICS), the same concerns surface repeatedly. Who owns an AI-generated output? Who is responsible if it is wrong? What data was the model trained on, and where does client information go once it has been processed? These are not abstract questions. Surveyors handle highly sensitive information, and there remains widespread misunderstanding about how Large Language Models work, particularly in relation to data retention, training, and risk. Until these issues are clearly addressed through governance and policy, many firms will remain cautious, preferring contained pilots to full-scale deployment.

Cultural and Organisational Inertia

Cultural and organisational factors further compound this caution. The UK property sector is, by nature, slow-moving and risk-averse. Revenues are long-term and relatively predictable (in particular for large REITs and housebuilders), margins are under pressure, and market conditions remain challenging. In such an environment, preserving cash often takes precedence over experimentation. Moreover, property firms tend to operate with relatively small headcounts compared to asset values or revenues, meaning that efficiency gains do not always translate into immediate reductions in FTE cost. This can make the return on investment harder to articulate in traditional terms.

Demographics also play a role. The average age within the sector continues to rise, and the industry struggles to attract younger, AI-native talent capable of driving change from within. This is not simply a question of skills, but of mindset and familiarity with digital tools as a default rather than an add-on.

The role of the RICS: Driving Responsible Adoption

The RICS is taking a global leadership role in preparing the surveying profession for the responsible adoption of AI. Recognising that AI is already transforming day-to-day practice, from valuation support and risk modeling to data extraction. The RICS has recently published a groundbreaking global professional standard on the responsible use of AI. This standard is designed not to stifle innovation but to support confident, ethical adoption by setting clear baseline expectations for competence, governance, and accountability among its 150,000 chartered surveyors worldwide.

The new conduct standard applies to all RICS members and regulated firms where AI outputs have a material impact on service delivery. Developed with key industry leadership, including Artefact co-chairing the working group, the framework concentrates on reinforcing professional judgement while augmenting expertise. The standard outlines critical requirements across five areas: establishing a baseline of AI literacy; strengthening practice management through governance and risk registers; introducing clear expectations for due diligence when procuring third-party AI tools; reinforcing professional judgement, scepticism, and transparency in relying on AI outputs; and setting accountability expectations for those involved in developing AI systems.

This proactive approach is crucial for managing the new professional risks introduced by AI, ensuring consistent practice, and protecting client trust. The RICS believes that by providing this shared framework, the profession can embrace innovation on a sound ethical and professional footing, attracting ‘AI Native’ talent and propelling the industry forward.

Crucially, the new standards are part of a wider, comprehensive RICS ecosystem to drive responsible AI adoption. This includes adapting the Assessments of Professional Competence (APCs), running practical training courses (such as the highly popular ‘Global Harnessing AI & Data in the Built Environment’), and publishing practical guidance documents. This concerted effort provides surveyors with a comprehensive toolkit, enabling them to deploy AI with greater confidence and integrity and ensuring the RICS remains relevant in the digital age.

Pragmatic Adoption – The key tenets

The inherent challenge lies in the human element: those within the property sector often display a reluctance, even an aversion, to embracing significant change. While regulatory inertia certainly plays a role, a more fundamental barrier exists in the prevailing attitudes. Conversations with senior property executives and surveyors frequently surface principles such as “if it isn’t broken, don’t fix it,” “no AI will replace my decades of deep contextual expertise,” or “our clients don’t pay us to use AI.” Although these perspectives contain a grain of truth, they more often signal a leadership preoccupied with preserving the status quo rather than pursuing transformative efficiency.

The initial flood of hyperbolic and often contradictory messaging surrounding AI proved counterproductive, breeding scepticism and confusion. However, we are now firmly navigating past the “Trough of Disillusionment” and beginning to ascend the “Slope of Enlightenment” (a recognisable pattern on the Gartner Hype Cycle). This transition means the focus must shift from ‘what if’ to ‘how.’ For the individual surveyor, this has created uncertainty about the practical starting point for adoption.

To foster genuine change, a dual approach combining ‘soft power’ and ‘hard power’ is essential.

From a ‘soft power’ perspective, organisations must visibly champion and encourage internal AI evangelists. This involves consistently carrying out engaging information and training sessions, establishing a continuous cycle of demonstrating early and tangible successes, openly sharing resources, documenting failures and findings, and ensuring the continuous upskilling of colleagues. Crucially, this cultural shift cannot take hold without clear, consistent, and active example-setting from the most senior leadership levels.

These cultural efforts must be robustly complemented by the ‘hard power’ of structural and operational reform. This necessitates irrevocably changing existing Ways of Working, mandating technical AI literacy certifications for key roles, redesigning organisational structures, refining operating models, and updating RACI matrices. The goal is to ensure that AI is no longer treated as a voluntary, “side-of-desk” activity, but is fully embedded into Business as Usual (BAU). This embedding must extend to budget allocations and performance metrics, treating AI deployment as a core business driver, not a peripheral technology project.

Above all, organisations must cultivate an environment that encourages experimentation, even if it leads to small-scale failure. This necessitates fostering honest, data-driven discussions about the practical utility of AI and the verifiable efficiency gains it delivers (or where it simply represents a ‘shiny tool’ with limited value). The most impactful applications of AI witnessed at Artefact are those which are seamlessly integrated into existing workflows. If the adoption of a tool requires significant friction for marginal gain, it will inevitably be abandoned en masse. Conversely, the greatest success often comes from those applications that are not necessarily visible but which quietly make professional life demonstrably easier by automating repetitive, manual, and ‘soul-destroying’ tasks. Being deliberate and surgical in identifying and deploying these high-impact, low-friction use cases is paramount to driving real, sustained adoption.

What Does The Future Look Like: Three-ish Time Horizons

Looking ahead, the opportunities for AI adoption within the UK’s property sector can be considered across three broad time horizons.

Short Term (Next 6 Months): Practical, Low-Risk Wins

The most compelling wins are practical and low-risk. Standardising instruction intake, automating first drafts of reports with clear evidence references, extracting structured checklists from leases and planning documents, and deploying internal knowledge tools that answer questions such as “how did we approach this last time?” all deliver immediate time savings. Notably, some of the largest gains are likely to be found in back-office functions such as HR, finance, and marketing, freeing surveyors to focus on client relationships and higher-value advisory work.

Medium Term (6 to 18 Months): Agentic Process Integration

The real value emerges from joining processes together. End-to-end transaction workflows powered by agentic AI, asset management platforms combining condition, compliance, energy, and capital expenditure data, and tools supporting retrofit and net-zero optimisation all fall into this category. Success here depends less on more advanced AI models and more on thoughtful process design and integration. All the foundational models have made huge strides recently in developing and standarding agentic capabilities so that they become child’s play to implement.

Longer Term (18 to 36 Months): Deeper Insight and New Services

Firms will increasingly turn their attention to deeper insight, including planning, geospatial, and climate risk analysis, portfolio-level digital twins, and new data-driven advisory services. However, these ambitions will only be realised where robust data governance and quality are already in place.

Even Longer Term (36+ Months): Existential Dread

Users of ‘X’, or even casual followers of general news, may feel a sense of overwhelming, almost existential dread. The pace of AI development is relentless. We are seeing humanoid robots replacing factory workers and even being deployed as soldiers, Self-Driving Cars finally hitting their stride, Neuralink promising to augment human capabilities, quantum computing inching toward reality, a massive acceleration in scientific research, and the widespread dominance of ‘AI slop’.

Yet, despite this ongoing cascade of discoveries and progress, the greatest current excitement is focused on the potential culmination of long-term efforts toward AGI (Artificial General Intelligence) — a far more contextually aware and complex AI system which could theoretically match, if not surpass, the human brain — and the profound societal shifts this implies. Will it truly boost productivity to such a degree that employment becomes optional and reliance on Universal Basic Income becomes a happy path forward?

The unprecedented capital being funneled into data centres, compute power, and infrastructure to achieve AGI is staggering. This investment will undoubtedly lead to significant market corrections in the near to mid-term. Return on Investment is likely to be slower than anticipated, and hype will probably outpace real-world impact. Nevertheless, it is poised to have a dramatic and unpredictable influence on our built environment and our interaction with it. Firms must remain agile and acutely cognizant of this reality if they are not only to survive but to thrive in this highly uncertain future.

The Non-Negotiables: Judgement, Trust, and Scrutiny

Throughout all of this, it is essential to remain clear-eyed about AI’s limitations. AI does not reduce professional responsibility; it concentrates it. Judgement cannot be delegated. While AI can assist analysis, it cannot replace the intuition and experience developed over decades, nor can it account for the emotional and often irrational human behaviour that shapes real estate markets. Surveyors remain legally responsible for their advice at all times, and that reality will not change.

Transparency with clients is equally critical. Where AI materially affects how a service is delivered, clients should be informed. This is not merely a compliance issue, but one of trust. Explainability and auditability are also essential. If a professional cannot explain where an output came from, what data was used, and what assumptions were made, defending that output in court may prove difficult.

Reliability risks should not be underestimated. AI systems can sound authoritative while being wrong, making review, sampling, and scrutiny indispensable. Finally, data protection and confidentiality remain non-negotiable. Prompts are records. Tools must be approved and assessed. Client data must be controlled. These are core elements of professional practice, not optional extras.
Ultimately, AI will not replace surveyors. What it will replace are manual searches, repetitive drafting, and low-value administrative work. Ultimately, the UK property sector’s path to AI adoption hinges not on adopting new technology, but on its willingness to fix old habits: embedding data quality, embracing the RICS governance framework, and consciously elevating human judgment above the tasks technology can now confidently handle.

About Artefact & Chris de Gruben, FRICS

Artefact is a global, full service Data and AI consultancy. We do everything from data strategy and AI governance, risk assessment, and compliance, to defining the art of the possible in AI and ML, to implementation and then to change management and adoption. Artefact leverages its deep expertise in both Property as well as Data Strategy & AI Governance to help firms to strategically embed AI in their operations responsibly and confidently.

Chris is a Senior Director at the UK Artefact office, leading the property team and managing all UK property clients. He has been a Chartered Surveyor for the past 15 years and continues to strongly advocate for the responsible use of AI within the UK property profession. Chris also serves as the Vice-Chair of the Professional Group on Valuation, sits on various AI expert working groups, is a core trainer for the RICS Academy, in particular the ‘Global Harnessing AI & Data in the Built Environment’ (which has proven to be the most popular RICS course so far). Chris is currently busy writing practical guidance for the use of AI for Chartered Valuers. Finally, he is a regular speaker at PropTech and Real Estate conferences globally.