A simple rule of thumb follows. Where outputs can be reliably checked, tasks lend themselves to automation. Where they cannot, AI is more usefully deployed as an augmenting tool.

McKinsey's report How agentic AI can reshape real estate’s operating model identifies four high-value domains for agentic AI in commercial real estate, these are :

• Maintenance and facilities: From dispatch to done automatically

• Leasing and renewals: Service, speed, and compliance

• Investing and asset management: Faster cycles and clearer judgment

• Construction and capital expenditures: Controlling complexity

Whilst the above four domains are clearly framed for real estate organisations, Corporate Real Estate professionals (those who manage their own employer's footprint) have every reason to follow the same route.

The final paragraph provides a perceptive insight -

The next phase is unlikely to be won by a series of disconnected pilots. It will be won by a small number of domain transformations designed for execution and trust. Organizations that move early can use agents to reduce handoffs, improve service, and accelerate decision cycles while building the governance and operating model that makes those gains durable. Done with discipline and risk controls, agentic AI can unlock new levels of efficiency, enhance customer service and experience, and fuel continuous improvement, delivering a true competitive advantage.

In other words, corporate real estate teams should focus pilot efforts on the redesign of entire workflow domains (end-to-end activity chains which leads to a defined outcome). For example, this could be an AI-native Carbon Tracker Agent capable of integrating BMS data, utility feeds and supply chain emissions in real time and automatically generating carbon compliance disclosures whilst continuously reviewing scope to reduce Scope 1/2 emissions.