AI in real estate: what changes for the Head of Real Estate in companies with their own sites
By Zoë Kschamer, Business Development Intern ·
Artificial intelligence (AI) in real estate mainly takes over the work that eats up time for companies with their own sites: gathering, reconciling and reporting data. Decisions, negotiations and alignment stay with the Head of Real Estate and tend to grow. The discipline is known as corporate real estate management (CREM). This article breaks the role down into seven task bundles and shows which of them AI automates or supports.
Most articles on AI in real estate deal with valuation, marketing and property management. This one takes the perspective of companies that run their own sites.
AI in real estate: where companies stand today
Corporate real estate (CRE) teams are using AI in more and more areas. For its 2025 Global Real Estate Technology Survey, JLL asked more than 1,000 decision-makers in 16 markets. In 2023, fewer than 5 per cent of occupiers planned to embed AI in their real estate processes. Two years later, 92 per cent had started pilots or planned to start them in 2025. Yet according to JLL, only 5 per cent had achieved most of their programme goals.
JLL's 2026 Future of Work Survey of more than 2,200 executives in 21 countries paints a similar picture. 78 per cent expect AI to change their portfolio strategy significantly over the next three to five years. Only 15 per cent have moved beyond exploration and initial deployment to using AI deliberately in their real estate operations. So far, 33 per cent are actively modelling how AI will affect their locations and asset types.
What AI takes over in real estate management: the role in seven task bundles
Asking whether AI in real estate management will replace a role keeps the discussion abstract. It becomes concrete once you split the role into task bundles and ask whether AI automates each one, supports it or leaves it with people.
| Task bundle | Effect of AI | What stays with people |
|---|---|---|
| Gathering and reconciling data (leases, headcount, utilisation, costs, energy) | largely automated | setting definitions, ensuring source quality |
| Reporting to the CFO, senior management and for sustainability reports | largely automated | deciding what senior management needs to see |
| Portfolio planning (consolidation, site network, reallocating space) | heavily supported | choosing and weighting the criteria |
| Business cases for investment and budgets | supported | owning the assumptions and the argument |
| Transactions (renewals, exits, negotiations with landlords and agents) | mostly human | negotiation, relationships, timing |
| Alignment with business units, HR, IT and the works council | human, growing in scope | more options mean more conversations |
| Managing service providers and facility management | supported | acting on performance data |
The table reflects an assessment of typical tasks and is not based on measurement. It still shows a pattern: the more a task consists of calculating and collecting, the more technology takes over. The more it depends on relationships and accountability, the more it stays with people.
AI in real estate portfolio management: the bottleneck moves from analysis to decision
Companies preparing a site consolidation usually compare only a few options, because each extra calculation costs time and consultancy budget. Once models handle these calculations, an extra option costs almost nothing. At the same time, the pace is picking up: according to JLL, space planning and location strategy are shifting from a once-a-decade project to a quarterly requirement.
It is tempting to conclude that more options lead to better decisions. Yet every additional option adds to the need for alignment. Take an example with fictitious figures: a company wants to merge three office sites in one region. Five options can be discussed in a single meeting with business units and the CFO. With 50 options that differ in target site, timing and floor area, that no longer works. If it is agreed upfront that costs must fall by at least 15 per cent and that no department's commute may increase by more than 20 minutes, most options drop out before anyone debates them. What remains are the few that meet both conditions. According to JLL's 2026 Future of Work Survey, the organisations further ahead do not have greater certainty about their future, but they develop agile planning that bridges the rapid cycles of AI with the longer-term commitments of real estate.
The shift has a second side. If the real estate team does not set these rules itself, other functions such as finance or procurement will. The role then remains one of supplying figures for decisions made elsewhere.
Five new tasks and the skills behind them
For the Head of Real Estate, this creates tasks that barely exist today:
- Setting the decision logic: defining the rules, weights and limits that models work with.
- Owning data quality: making sure data from HR, finance and facility management (FM) is reliable, even when other functions own those systems.
- Challenging results: recognising when a recommendation rests on a wrong assumption.
- Planning continuously: reviewing the portfolio every quarter, regardless of the next lease end.
- Explaining results: presenting them in a way the CFO, the works council and employees can follow.
None of this requires coding. What counts is data literacy, meaning knowledge of sources, definitions and gaps, and the ability to translate business strategy into criteria and weights. Add to that comfort with ranges and trade-offs, and the language of the CFO, such as cost per employee or carbon per square metre. This is currently the biggest hurdle: in the 2026 Future of Work Survey, gaps in AI, analytics and emerging technology skills overtook budget as the main barrier to real estate transformation for the first time in 15 years.
What needs to be in place first: data, ownership, criteria and co-determination
Tools for corporate real estate AI are usually quicker to introduce than the foundations. The starting point is a shared data set: leases, floor area, headcount, utilisation, costs and energy in one place, with agreed definitions. This covers what counts as a workplace, when a desk counts as occupied and which floor area the metrics refer to. Our article on real estate portfolio KPIs explains which metrics follow from this and how to calculate them.
Because the data comes from HR, finance, FM and IT, everyone needs to know what they update and how often. The decision criteria from the section on portfolio management are written down and signed off by the CFO or senior management.
In Germany, works councils have co-determination rights under section 87(1) no. 6 of the Works Constitution Act (Betriebsverfassungsgesetz, BetrVG) for technical equipment designed to monitor employees' behaviour or performance. Whether utilisation data falls under this depends, among other things, on whether it can be linked to individuals. This needs legal review in each case, as do the requirements of the GDPR. Because a works agreement takes time to negotiate, the conversation with the works council belongs at the start.
Roles within the team shift. Collecting figures takes less time, while data ownership, transactions and stakeholder alignment grow. External advisers are needed mainly where experience in valuation and negotiation matters.
The first 90 days: start with a decision that is already due
A good starting point is a decision that is coming up anyway, such as a lease renewal. In the first four weeks, the team records data sources and gaps, along with every lease event in the next 24 months. In weeks five to eight, the Head of Real Estate agrees definitions and criteria with the CFO and opens the conversation with the works council.
In weeks nine to twelve, the decision is run through scenarios, starting with costs, floor area and remaining lease terms. Utilisation data is added once the agreement with the works council is in place. The result shows internally what the new way of working delivers and forms the basis for the next decision.
With spaciv's portfolio strategy, you can combine utilisation data from your existing systems with cost, space and site data based on consistent definitions, and run scenarios for your portfolio before a decision is due.
Frequently asked questions
Will AI replace the Head of Real Estate?
No. It mainly changes what the role spends its time on. Respondents to JLL's 2026 Future of Work Survey also largely expect AI to redesign roles, and most anticipate growing workforces.
What is AI used for in real estate?
At companies that use their own sites, pilot projects focus on three areas, according to JLL. In data workflows, the aim is to standardise data, integrate sources and automate reports. In portfolio optimisation, AI supports portfolio analysis, strategy and capital planning. In energy management, it analyses consumption, helps plan decarbonisation and controls heating, ventilation and air conditioning. Extracting key data from leases is widely seen as an obvious use for generative AI, but it is not where the pilot projects are concentrated.
Do you need new software to use AI in portfolio planning?
Not necessarily. Many companies already work with systems such as CAFM (computer-aided facility management) or IWMS (integrated workplace management system). The real question is whether their data can be combined with other sources. With CAFM+, spaciv builds on existing CAFM data without replacing the system. According to JLL, more than 60 per cent of companies first need to fix basic issues in their existing systems, such as duplicated functionality or dormant software, before they can make full use of AI.
Sources
Retrieved on 03.11.2026, Future of Work Survey figures rechecked on 11.11.2026. Sources available in German only are marked.
