Artificial intelligence (AI) has the potential to improve business efficiencies, disrupt employment markets and optimise tenant experience. In this report, we examine the potential impacts of AI on the office sector
AI’s growth, and where are we now?
What is AI?
AI is a technology that enables computers and machines to simulate human intelligence to solve problems and improve productivity. Generative AI uses algorithms that allow systems to make predictions based on learning from existing data to create new data. However, there remains confusion among the AI buzz as to whether businesses are implementing a new form of technology or if they are actually employing AI/machine learning (ML).
Public and private institutions are investing in AI to drive business growth, productivity and competitiveness. Investors have recognised AI as a future growth sector, with over $90bn of venture capital funding raised by AI and ML companies globally during 2023.
Why is AI experiencing such growth?
AI adoption is increasing across European businesses, although remains in its infancy. On average, 8% of European companies had adopted at least one application of AI during 2023, with the UK and northern Europe leading the way. Of those businesses currently using or planning to use AI applications, the most common reasons for doing so were 1) to improve cybersecurity (35%) and 2) to create efficiencies (35%). Clearly, the benefits of enhanced internal knowledge transfer through AI, or ‘Corporate AI’, will act as a huge efficiency gain across larger companies.
Deloitte’s UK CFO Survey Q2 2024 indicates that generative AI was perceived as the biggest internal risk to businesses, with primary concerns focussed around a shortage of technical skills among employees, and risk of adoption.
Impact on office-based jobs
Internet 2.0
AI has been referred to by commentators as the biggest digital change since the emergence of the internet. However, rising internet adoption within UK households through the late 1990s and early 2000s did not have a significant impact on office-based employment or office demand. In fact, Inner London’s office-based employment has more than doubled since 1998, over which time, the average annual London City office take-up has increased by 66%.
Instead, productivity growth and office-based employment growth have traditionally been the drivers of office demand. Goldman Sachs estimates that 60% of workers today are employed in jobs that did not exist in 1940 – as technology displaces obsolete roles, new roles are created.
Less automation, more augmentation
The speed of adoption of AI is increasing, and more focus is shifting to the impact that AI will have on office-based jobs for the first time.
AI is generally expected to be most effective in terms of reducing the time spent compiling/consolidating information, freeing up human labour for more value-add work, including:
- implementation/delivery of tasks
- tasks requiring more creative/collaborative solutions
- business-generating output
Clearly, employers will seek an augmented skillset from employees to best utilise AI applications, with Oxford Economics forecasting that 12% of workplace tasks will be fully automatable by 2032. However, this does not correspond with all jobs, as each job faces different degrees of exposure to AI based on the composition of tasks. The fastest-growing job roles are likely to be within technology, digitalisation and sustainability, with a decline in the number of roles that are solely clerical/secretarial focussed.
Over the longer term, we expect to continue to see a structural shift towards service sector job growth and office-using employment – Oxford Economics forecasts eurozone office-based employment to increase by over one million jobs over the next ten years. As AI adoption increases, we expect job churn to increase over the short/medium term, as employers decide which skillsets they are seeking and workers take time to reskill.
Governments and businesses will, therefore, upskill and retrain staff in order for them to utilise rather than compete with AI. Analytical and creative thinking, AI/big data and leadership are among the key skills that employers are seeking to upskill over the next five years, according to the World Economic Forum (WEF).
How are different occupiers adapting their requirements?
AI-related companies are expanding rapidly across European cities – among the quickest is Central London, upsizing at rates of up to ten times existing floorspace, as they compete to hire the best talent.
Across Europe this year, Shoper Group signed for 600 sq m as a development centre for their AI operations in Poznan, Poland; US global semiconductor company AMD doubled its footprint, signing for 4,200 sq m at Cambridge Science Park, UK, and AI company Amadeus signed for 500 sq m in Milan, Italy. Similar to life sciences, AI remains one of the industries where physical clusters with access to knowledge and expertise remain most important among occupiers.
On a broader scale, we are seeing a new wave of demand coming from the service sector to expand their digital functions, with a particular focus on legal, data and cybersecurity in order to ensure compliance and security. Big tech companies are using ‘red teams‘ of hackers to test their AI models before they are released more widely.
Companies want to use AI but are not sure what they want to use it for, or how best to implement it in their operations. We expect companies will seek more flexible lease terms as they plan for AI and scale headcount accordingly.
Some tech employers have previously used AI as a reason for cutting jobs, although are rehiring staff in different roles, which has helped to reset wage scales.
The legal sector is one of the areas where we are seeing companies plan furthest ahead – many paralegals are using AI applications for document reading/summarising, for example, and paralegals are responsible for approving the AI-generated output rather than compiling it. Companies are still some way away from fully trusting the output but are using it to improve efficiencies in more time-intensive, low-skilled tasks.
How is AI being used to improve occupier experience?
From a building management perspective, landlords are employing more technology in order to improve building efficiency; however, AI usage generally remains in its infancy. Early-stage examples include:
- Building managers are using AI systems to a) identify any erroneous energy usage data, b) scrape building and legal documents to extract key information and automate actions, c) quickly access building technical information to support maintenance work etc.
- Chatbots are helping buildings managers by gathering and collating building faults from tenants, so building managers can implement solutions more quickly and improve customer experience.
- Occupiers are utilising building management systems to direct employees to where they should sit in the office, in order to reduce energy usage on off-peak days. This more effectively manages the lift/ elevator capacity and lighting/heating costs on unutilised floors.
- Workplace advisory teams are increasingly able to provide real-time office design visuals as an indication to advise occupiers on fitout and improve speed of delivery. As AI will shift human labour to more collaborative, value-add tasks, over the longer term, we expect workplaces to feature more breakout areas and meeting rooms.
Challenges for the growth of AI
AI’s huge potential does not come without its challenges. We outline the barriers to growth
Governance/Social Reliability – One of the largest concerns of AI-generated output is whether it can be fully trusted. Companies may seek to improve efficiencies through AI, but they are some distance from being comfortable to fully automate processes without human approval, given the data can be pulled from open sources.
Data security – Despite potential efficiency gains, companies also remain cautious over the security of the data they have shared and want to ensure this is kept ring-fenced from public use.
Inclusion and diversity – Does AI make diversity and social inclusion easier or harder?
Job displacement – The WEF’s research indicates a displacement of 2% of jobs in the next five years, and governments and businesses have a responsibility to ensure employees are upskilled and reskilled. The EU’s AI Skills Strategy for Europe provides a roadmap to ensure Europe can meet the growing demand for AI-enabled professionals. The focus will be on providing current and professional roles across four domains: AI practitioners, AI management and support, organisational decision-makers, and policymakers.
Grid capacity and sustainability concerns
One of the barriers to the growth of AI is the size of the grid capacity to keep up with AI development. The WEF estimates that the computational power required for sustaining AI’s rise is doubling roughly every 100 days, and rising AI demand has to compete with other energy-intensive industries for power.
Cornell University’s study also indicates that a generative AI system requires 33 times more energy to complete a task than with traditional software, resulting in both higher carbon emissions and water usage, placing strain on electricity grids. As such, big tech companies’ Scope 3 emissions have increased by between circa 50–65% over the last three years.
Regulatory headwinds
As part of its digital strategy, the EU’s AI Act wants to regulate AI to ensure better conditions for the development and use of technology – to ensure AI systems are safe, transparent and traceable. Generative AI will have to comply with transparency requirements, including disclosing that the content has been generated by AI, designing the model to prevent it from generating illegal content, and publishing summaries of copyrighted data for training, to ensure transparency to end users.
Conclusion
We are still at an early stage of AI adoption, and there remain governmental, environmental and regulatory challenges to further growth.
Businesses are far from fully devolving responsibility to AI in the short to medium term. Many jobs will need to adapt to incorporate and utilise AI, with jobs both lost and created as governments and businesses focus on upskilling and reskilling workers. We believe that the medium-long-term structural trend towards service sector employment growth will outweigh the number of office-based roles automated through AI.
The divergence between sectors will be varied, with technology and programming-based roles observing large increases and administrative and clerical roles declining. Though, this is nothing new – as with any new technological advancement, job displacement and retraining have always been the case and governments and businesses will play a large part in the reskilling of workers.
Occupiers will also demand more from their building experience, particularly with improved efficiencies around heating, ventilation and air conditioning (HVAC) as a way to reduce energy usage. For office occupiers who are faster adopters of AI, we expect workplaces to be more focussed towards in-person collaboration, as companies continue to adjust to their hybrid working models.