Artificial intelligence is rapidly finding its way into commercial property, with property management often cited as one of the areas most likely to benefit. From automating routine tasks to improving decision-making, the potential is clear. However, as discussed by Josh Panknin – including in his recent sessions at MIPIM – the reality is more grounded. AI is not a silver bullet, and in a sector as complex and fragmented as property management, its value depends entirely on how it is applied.
At its best, AI can enhance efficiency and provide sharper insights. Property management is inherently operational: dealing with maintenance, service charge budgets, tenant communication, compliance, and financial oversight. Much of this relies on large volumes of data, often spread across systems and formats. AI offers the ability to consolidate and interpret this data more effectively.
For example, predictive analytics can identify maintenance issues before they become costly problems. Analysing patterns in repair history, building age, and usage data allows managers to move from reactive to proactive maintenance. Similarly, AI can assist with service charge forecasting, flagging anomalies or trends that may otherwise go unnoticed until year-end reconciliation. On the tenant side, automated communication tools can streamline responses to common queries, improving response times without increasing workload.
There is also a growing role for AI in portfolio level decision making. By analysing performance across assets, managers can identify underperforming properties, forecast income risk, and support strategic decisions around lease events or capital expenditure. In theory, this creates a more informed, data-driven approach to asset management.
However, this is where Panknin’s perspective becomes particularly relevant. His work consistently emphasises that AI in real estate only works when it is grounded in a deep understanding of the underlying processes. Property management is not a clean dataset, it is shaped by lease terms, tenant behaviour, building condition, and local market dynamics. These variables are often inconsistent and difficult to standardise.
One of the primary risks, therefore, is poor implementation. Many AI initiatives fail not because the technology is flawed, but because the problem has not been properly defined. In property management, this often manifests as tools that produce outputs without context. For instance, an algorithm may flag a property as underperforming without recognising that a short-term vacancy is part of a wider asset strategy. Without human interpretation, these insights can be misleading.
Data quality is another critical issue. Much of the information used in property management is incomplete, outdated, or inconsistently recorded. AI models rely on clean, structured data; without it, outputs become unreliable. This is particularly relevant in secondary stock, where records are often less robust and buildings have evolved over time. Introducing AI into such environments without addressing data quality can create a false sense of accuracy.
There is also a risk of over-automation. While AI can handle repetitive tasks, property management remains a relationship-driven business. Tenant satisfaction, negotiation, and judgement calls cannot be fully automated. Over-reliance on technology risks eroding the human element that underpins effective management.
Perhaps the most important takeaway from Panknin’s work is that AI should be viewed as a tool, not a solution. Its role is to support better decision-making, not replace it. The most successful applications are likely to be targeted and practical, automating specific processes, improving data visibility, and enhancing existing workflows.
For property management companies like Hootons, the opportunity lies in adopting AI selectively and with clear intent. I believe, those who take this approach are far more likely to see meaningful benefits, while avoiding the common pitfalls.
In essence, AI will undoubtedly play a role in the future of property management. The challenge is ensuring that its adoption is grounded in reality, not driven by hype.
Top image: panel at MIPIM 2026 (this article does not reflect the content of the panel).
