AI and offices

The impact of AI on the office sector will depend upon whether AI leads to job growth or job displacement and whether occupier demand will increase or decrease as a result. And this must be layered over the structural changes the sector is already seeing due to the impact of hybrid working. Moreover, different occupations will have different levels of exposure to generative AI, based on the task and expected rate of adoption.

Productivity enhancing AI should increase the profitability of corporates, and hence, their ability and willingness to pay higher rents. Similarly, sectors which see rising job numbers due to AI should see demand for space rise. By contrast, those sectors which are disrupted by AI and see profitability squeezed and workers displaced are most at risk of rental declines. This would be akin to the structural shocks which have impacted the retail and office markets in recent years.

Initially we expect there to be a greater focus on the automation of routine tasks, these would typically be back-office roles which easily lend themselves to AI. In many cases, corporates have outsourced these types of roles to lower cost locations such as India and Eastern Europe, hinting that the first wave of AI automation is less likely to impact core CBD locations in the advanced economies, while back-office hubs will likely be more exposed.

According to Oxford Economics, in the US office is the property sector most exposed to the impact of generative AI, followed by life sciences and manufacturing. These sectors will likely feel a negative impact on occupier demand. In the case of life sciences, productivity enhancements from AI need to be considered against overall growth in the sector. Office-using occupations have greater potential for generative AI to automate tasks and are likely to have a higher adoption rate. This adds to the headwinds that the global office sector is already facing from hybrid working, working age population decline and decarbonization-related capex.

As the adoption of generative AI to assist and automate workplace tasks increases, some office workers will be displaced. By 2032, Oxford Economics estimates that over 9% of the current US workforce will be displaced by generative AI. The affected workers are predominantly in the office-using sectors such as information, professional, scientific and technical services. This will likely present a major headwind to future space demand for offices.

The US economy is one of the most dynamic in the world though and has successfully adopted and embraced change in the past. However, there will likely be short-term pain points. Around 80% of the workers expected to be displaced will likely find new roles, though the lag in time for workers to be re-employed hits productivity for the economy as a whole.¹  Office and administrative support and sales roles are highly exposed to generative AI and likely to face frictions when trying to move to alternative positions.

On the positive side, not only will AI model developers such as OpenAI represent a growing occupier segment for offices, but so will companies involved in semi-conductor hardware, cloud computing platforms and application development which are also part of the AI ecosystem. According to Stanford University, in 2023, the area receiving the most AI investment globally was AI infrastructure/research/governance, followed by natural language processing (NLP) and customer support; data management and processing; and medical and healthcare.²  Furthermore, AI can improve the efficiency of sectors that are operationally intensive like healthcare, hotels and senior housing.

Source:
¹ Oxford Economics, March 2024
² Artificial Intelligence Index Report 2024, Stanford University