The substitution effect: AI and the labour market

The substitution effect: AI and the labour market

At Columbia Threadneedle Investments, we believe firms which recognise and manage their human capital effectively will outperform in the longer term. To help us assess how well an investee firm is doing this we created the “Five ‘S’ framework of human capital management”1. This sets out five factors a firm may consider when managing human capital (Figure 1).

In this Thematic Insight we focus on ”substitution” – that is the opportunities firms have to substitute human capital for automation, digitalisation and, increasingly, artificial intelligence (see Definitions box) to help lower operating costs, increase productivity and/or to address talent and skills shortages.

Figure 1: the Five ‘S’ framework of human capital

Figure one the Five S framework of human capital
Source: Columbia Threadneedle Investments, 2023

Jobs lost, jobs gained, jobs changed

Just as the world of work was getting back on a more familiar footing following the seismic impact of the Covid-19 pandemic, along came the emergence of generative AI with the potential to further disrupt labour markets – including the companies in which Columbia Threadneedle invests.

According to a report by Goldman Sachs, generative AI could “expose” the equivalent of 300 million jobs globally to automation.2 While there are very few occupations – perhaps less than 5% – which are fully automatable, six out of 10 could have around 30% of activities automated.3
Clerical and administrative jobs are the most exposed to technological advancement. In fact, we have already seen IBM4 and BT5 announce they will reduce back office and customer service roles and replace them with technology including AI. On the flip side, demand from a broad range of industries for software developers and cybersecurity specialists is increasing significantly.6

The quantum of global jobs lost through firms substituting labour for new technologies, or indeed created by firms finding new ways of manufacturing goods and services, is unknown. What is apparent, however, is that a significant number of jobs done today will change.

History lesson

History gives some indication of the long-term impact of technological advancement – be it Henry Ford’s automobile assembly line of the 1900s10 or more recently the introduction of personal computers in the 1970s.11 Firms have adapted to new technologies, improved productivity and revenues, and created new roles for humans to do. Indeed, despite the technological advancements of the past century many advanced economies are currently experiencing near full employment.12
What is potentially different this time is the breadth of the impact of technological advancements across a spectrum of roles13 and across the many sectors in which Columbia Threadneedle invests, including healthcare, legal services, consumer, media, manufacturing and education. New technologies including generative AI have the potential to automate repetitive and routine tasks and free up employees’ time to focus on more complex and creative tasks.14

To maximise the opportunities AI and automation can provide, companies will not only need to invest in technology but manage their human capital effectively – reskilling and retraining existing employees, and promoting strong labour relations and supportive cultures. Alongside this they will need to attract new employees with distinct skillsets from their traditional hires.

AI sector opportunities and the implications for human capital management

Healthcare: supporting an ageing population and talent shortage

AI is already incorporated in healthcare products and services such as imaging and diagnostics, assisting with surgeries, drug development and supporting operational efficiency in hospitals. This is beneficial to a global health system under considerable pressure due to ageing populations and constraints to government spending. Wider adoption of AI (Figure 2) has the potential to reshape the sector by improving patient outcomes and driving efficiency gains. It is estimated it could lead to savings of 5%-10% in US healthcare spending (roughly $200 billion–$360 billion annually.15

However, talent shortages persist across the broad range of healthcare practitioners from healthcare assistants to biopharma lab technicians. According to LabioTech, in the biopharma sector alone there are in excess of 800,000 employees but more than 60,000 job vacancies, which indicates a labour shortage of approximately 8%.16  Projections show that job opportunities in the life, physical and social sciences sectors will grow by 7% by 2028.17 While AI may mitigate some of these shortages – virtual wards can support some aspects of remote healthcare provision – the role of labour within hospitals and within R&D will remain prominent. Shortages will need to be addressed for companies to meet their growth strategies.

Columbia Threadneedle will engage with healthcare companies in understanding their approaches to talent management and seek to understand how they are planning to manage the risks associated with AI within healthcare – for example, around data breaches.

Projections show that job opportunities in the life, physical and social sciences sectors will grow by 7% by 2028

Utilities: powering the future

Utility companies are vital to global economies, generating, transmitting and distributing the gas, electricity and water required to run businesses and households. However, the transformation required by these firms as the world shifts to a more sustainable energy system will be significant.

To support the transition to a net zero world, utility companies are looking to technology to improve efficiency, reduce costs and enhance service delivery. Drones can support defect detection and predictive maintenance in pipes, wiring or machinery; AI can enhance the consumer experience by offering dynamic lower prices when there is excess capacity; and generative AI could help tackle the challenge of integrating variable and unpredictable renewable assets into the energy generation mix through analysis of weather patterns and scenario analysis.

However, building these capabilities in the breadth and depth required by utility companies will involve significant talent acquisition and reskilling of workforces. At the Columbia Threadneedle Investment Energy Transition Conference in May 2023, one of the consistent themes from companies critical to the energy transition was labour as a pinch point, particularly in relation to grid modernisation.18 This could have implications around, for example, the potential of electric vehicles and heat pumps to help decarbonise the planet.

Columbia Threadneedle will continue to engage with firms which are key in decarbonising the energy system to better understand their talent management processes, as well as the opportunities they are exploring in new technologies to help the transition.

Catering companies: increased efficiencies and managing
recruitment challenges

Catering companies are labour intensive businesses. Last year, Compass reported that it recruited 110,000 individuals in North America alone.19 Labour as a percentage of revenue for the two largest catering companies in Europe, Sodexo and Compass Group, are around 47% and 49% respectively,20 while voluntary turnover of staff is high at 29%21 and 35%.22 Both companies are experiencing strong demand as firms look to outsource their culinary staff offerings to focus on their core business, which is putting further pressure on staffing demands in tight labour markets.

While the need for onsite human employees will persist, Compass and Sodexo are both focused on improving efficiencies and investigating opportunities around automation to mitigate some of the labour supply pressures they are facing globally. Examples from Compass include the North Bar Tap + Go frictionless kiosk at Leicester City Football Club,23 which removes the need for cashiers, and fully robotic kitchens – van-sized kitchens offering fresh meals 24-7 – in a healthcare setting.24

Catering companies are also using AI to leverage their data across their operations. AI technologies have allowed catering companies to better predict peaks in customer footfall, are used within CRM software to direct sales efforts towards the most valuable opportunities and help digitalise financial operations and administrative tasks. All of which have improved productivity and helped manage costs.

Going forward, generative AI may support sales teams communicate with potential clients and promote ways to upsell to existing clients. Generative AI may also help design the most efficient layouts of canteens and create pitch documents.

Columbia Threadneedle will continue to engage with catering companies to better understand the opportunities generative AI, automation and other technologies can offer to offset the risks inherent in a labour-intensive business. Likewise, we will explore the potential risks of generative AI on those sectors which use catering companies to provide food services, such as higher education settings.

Engagement case study: Relx25 – risk or revenue opportunity?


The release in November 2022 of ChatGPT, a large language model owned by OpenAI and the fastest platform to a million users,26 raised market concerns regarding the possible impact of the technology on sectors most exposed to it.

As part of our regular engagement with senior management teams, Columbia Threadneedle Investments met with European-based global business information services firm Relx. It provides information-based analytics and decision-making tools to help clients, such as those in insurance, legal or medical, to manage risks.
We wanted to understand how management saw the perceived risks and opportunities associated with generative AI, and how the company would ensure it has the appropriate human capital strategy, such as skilled labour, to deal with this transition.
Management outlined how their businesses have been using AI for more than a decade to support clients in analysing and making decisions. Insightfully, management believe it is not just access to the tech itself that gives the company a competitive edge, rather the proprietary data set and deep customer knowledge. As they put it, “the power is having all three”.
The CEO explained that while some of the data they use is publicly available, it has been collected over decades and often first hand (transcribing from within court houses) before being checked, stored and formatted in a searchable way in their own databases. This data is then overlayed with expert opinion and classified and mixed with proprietary data. The resulting database is, they conclude, fully proprietary. To replicate it would be extremely difficult, time-consuming and expensive.
Relx are not complacent, however, and are exploring ways they can improve their offering using generative AI. They have a product in trial that can respond to more conversational question searches, summarise cases and create draft client letters. They are also experimenting with numerous LLM models and see generative AI as an incremental positive for their long-term growth strategy rather than a significant threat.
In relation to human capital and attracting talent, Relx believe that the most important element of human capital is motivating staff through the “purpose” of the company – Relx products help their clients: they improve scientific study or medical outcomes, improve point of law and prevent fraud – this resonates with employees, and they are able to attract and retain good staff.


Columbia Threadneedle will continue to assess the impact of AI and other substitution opportunities on the growth strategies of the firms we invest in. Using the Five S framework we will also evaluate the associated implications for firms’ human capital management.


Automation is a type of software that follows pre-programmed rules. It is the application of technology, programs, robotics or processes to achieve outcomes with minimal human input. There are different types of automation including “basic”, “process” and “intelligent” automation. The latter incorporates AI and machine learning capabilities, allowing machines to “learn” and take better decisions and actions based on data from situations it has previously encountered and analysed. An example would be a virtual customer assistant.7
Digitisation and digitalisation refers to taking analogue information and converting it into a digital format.8 Digitalisation is the use of digital technologies to change a business model and potentially provide new revenue opportunities.9
Artificial intelligence (AI) is the broad field which refers to the use of technologies to build machines and computers with the ability to mimic cognitive functions associated with human intelligence. These include being able to see, understand and respond to spoken or written language, analyse data and make recommendations, and more. It is a set of technologies implemented in a system to enable it to reason, learn and act to solve a complex problem. Common applications include chatbots to replace customer service on numerous company websites and face recognition on smart phones.
Machine Learning (ML) is a subset of AI that automatically enables a machine or system to learn and improve from experience. Instead of explicit programming, ML uses algorithms to analyse large amounts of data, learn from the insights, and make informed decisions. ML algorithms improve over time as they are trained – that is, exposed to more data. ML models are the output, or what the programmes learn from running the algorithm on training data. The more data used, the better the model will get. wa
ML is a discipline of AI that provides machines with the capacity to automatically learn from data and previous experiences. They do this by identifying patterns to generate predictions for new processes with minimal human intervention. ML is applied in several situations where it is impossible to apply strict algorithms.
Common applications of ML include image recognition used in healthcare settings which can support the detection of anomalies; e-commerce product recommendations; the detection of spam emails; and, despite the name, algorithmic trading.
Generative AI is an advanced branch of AI that utilises machine learning techniques to generate original content such as images, text, audio and video. Unlike traditional ML, which focuses on mapping input to output, generative models aim to produce novel and realistic outputs based on the patterns and information present in the training data. Dall-E or Chat GPT-4 are examples of generative AI.
1 Read more in our previous viewpoint Managing the human capital transition, August 2023
2 Goldman Sachs: The potentially Large Effects of Artificial Intelligence on Economic Growth (Briggs/Kodnani), 26 March 2023
3 McKinsey & Company, Jobs lost, jobs gained: workforce transitions in a time of automation, December 2017
4 Techmonitor, IBM could replace 7,800 back-office workers with AI, 2 May 2023
5 BBC, BT to cut 55,000 jobs with up to a fifth replaced by AI, 18 May 2023
6, Dice Reports Strong Tech Hiring Results Across Key Industry Sectors, 8 August 2023
7 IBM,
8 Forbes: Digitization, Digitalization, And Digital Transformation: Confuse Them At Your Peril, 29 April 2018
9 Forbes: Digitization, Digitalization, And Digital Transformation: Confuse Them At Your Peril, 29 April 2018
10 Ford, The moving assembly line and the five-dollar workday, 2023
11 McKinsey & Company, Jobs lost, jobs gained: workforce transitions in a time of automation, December 2017
12 UN, World Economic Situation and Prospects: October 2023 Briefing, No. 176, 2 October 2023
13 McKinsey & Company, Jobs lost, jobs gained: workforce transitions in a time of automation, December 2017
14 LinkedIn, Revolutionizing the Workplace: How Generative AI, Including ChatGPT, is Changing the Workforce and Boosting Efficiency, 10 July 2023
15 Healthcare Dive, Artificial intelligence could save healthcare industry $360B a year, 26 January 2023
16, Recruiting and Retaining Talent Is the Biggest Challenge Facing the Pharmaceutical Industry, 16 June 2022
17 LabioTech EU, Tackling the skilled labor shortage in biopharma manufacturing, 27 September 2023
18 Columbia Threadneedle Investments, The energy transition – transformative on a global scale, August 2023
19 Compass Group Careers,, 2022
20 Columbia Threadneedle Investments’ analysis, 2023
21 Bloomberg/Sodexo Fiscal 2022 Universal Registration Document, 2022
22 Compass Group, GRI Index 2022, 2023
23 LCFC, King Power Stadium First in Europe To Introduce Frictionless Kiosk, 6 August 2022
24 Food On Demand, Compass Group, Rowok announces 3-Location Robotics Pilot, 20 July 2022
25 The mention of specific companies is not a recommendation
26 Reuters, ChatGPT sets record for fastest-growing user base, February 2023
5 January 2024
Sally Springer
Sally Springer
Senior Thematic Research Analyst, Global Research
Share article
Share on linkedin
Share on email
Key topics
Related topics
Listen on stitcher badge
Share on twitter
Share on linkedin
Share on email
Key topics
Related topics


The substitution effect: AI and the labour market

Related Insights

22 May 2024

Joanna Tano

Head of Research, Europe, Real Estate (EMEA)

UK Real estate – Overview Q1 2024

Total returns for the UK commercial property market turned positive in Q1 2024.
Read time - 3 min
21 May 2024

Pauline Grange

Portfolio Manager, Global Equities

Water in crisis – searching for solutions

With too much, too little or too toxic water the world is facing a water crisis. We explore key issues and challenges before highlighting some of the companies promoting better water management.
Read time - 3 min
20 May 2024

Steven Bell

Chief Economist, EMEA

Inflation falls: how far will interest rates follow?

Inflation is finally falling in the US, UK and Europe, but will we see greater dispersion in central bank actions?
Watch time - 4 min

Risk Disclaimer

For use by professional clients and/or equivalent investor types in your jurisdiction (not to be used with or passed on to retail clients). For marketing purposes.


This document is intended for informational purposes only and should not be considered representative of any particular investment. This should not be considered an offer or solicitation to buy or sell any securities or other financial instruments, or to provide investment advice or services. Investing involves risk including the risk of loss of principal. Your capital is at risk. Market risk may affect a single issuer, sector of the economy, industry or the market as a whole. The value of investments is not guaranteed, and therefore an investor may not get back the amount invested. International investing involves certain risks and volatility due to potential political, economic or currency fluctuations and different financial and accounting standards. The securities included herein are for illustrative purposes only, subject to change and should not be construed as a recommendation to buy or sell. Securities discussed may or may not prove profitable. The views expressed are as of the date given, may change as market or other conditions change and may differ from views expressed by other Columbia Threadneedle Investments (Columbia Threadneedle) associates or affiliates. Actual investments or investment decisions made by Columbia Threadneedle and its affiliates, whether for its own account or on behalf of clients, may not necessarily reflect the views expressed. This information is not intended to provide investment advice and does not take into consideration individual investor circumstances. Investment decisions should always be made based on an investor’s specific financial needs, objectives, goals, time horizon and risk tolerance. Asset classes described may not be suitable for all investors. Past performance does not guarantee future results, and no forecast should be considered a guarantee either. Information and opinions provided by third parties have been obtained from sources believed to be reliable, but accuracy and completeness cannot be guaranteed. This document and its contents have not been reviewed by any regulatory authority.


In Australia: Issued by Threadneedle Investments Singapore (Pte.) Limited [“TIS”], ARBN 600 027 414. TIS is exempt from the requirement to hold an Australian financial services licence under the Corporations Act 2001 (Cth) and relies on Class Order 03/1102 in respect of the financial services it provides to wholesale clients in Australia. This document should only be distributed in Australia to “wholesale clients” as defined in Section 761G of the Corporations Act. TIS is regulated in Singapore (Registration number: 201101559W) by the Monetary Authority of Singapore under the Securities and Futures Act (Chapter 289), which differ from Australian laws.


In Singapore: Issued by Threadneedle Investments Singapore (Pte.) Limited, 3 Killiney Road, #07-07, Winsland House 1, Singapore 239519, which is regulated in Singapore by the Monetary Authority of Singapore under the Securities and Futures Act (Chapter 289). Registration number: 201101559W. This advertisement has not been reviewed by the Monetary Authority of Singapore.


In Hong Kong: Issued by Threadneedle Portfolio Services Hong Kong Limited 天利投資管理香港有限公司. Unit 3004, Two Exchange Square, 8 Connaught Place, Hong Kong, which is licensed by the Securities and Futures Commission (“SFC”) to conduct Type 1 regulated activities (CE:AQA779). Registered in Hong Kong under the Companies Ordinance (Chapter 622), No. 1173058.


In Japan: Issued by Columbia Threadneedle Investments Japan Co., Ltd. Financial Instruments Business Operator, The Director-General of Kanto Local Finance Bureau (FIBO) No.3281, and a member of Japan Investment Advisers Association and Type II Financial Instruments Firms Association.


In the UK: Issued by Threadneedle Asset Management Limited, No. 573204 and/or Columbia Threadneedle Management Limited, No. 517895, both registered in England and Wales and authorised and regulated in the UK by the Financial Conduct Authority.


In the EEA: Issued by Threadneedle Management Luxembourg S.A., registered with the Registre de Commerce et des Sociétés (Luxembourg), No. B 110242 and/or Columbia Threadneedle Netherlands B.V., regulated by the Dutch Authority for the Financial Markets (AFM), registered No. 08068841.


In Switzerland: Issued by Threadneedle Portfolio Services AG, an unregulated Swiss firm or Columbia Threadneedle Management (Swiss) GmbH, acting as representative office of Columbia Threadneedle Management Limited, authorised and regulated by the Swiss Financial Market Supervisory Authority (FINMA).


In the Middle East: This document is distributed by Columbia Threadneedle Investments (ME) Limited, which is regulated by the Dubai Financial Services Authority (DFSA). For Distributors: This document is intended to provide distributors with information about Group products and services and is not for further distribution. For Institutional Clients: The information in this document is not intended as financial advice and is only intended for persons with appropriate investment knowledge and who meet the regulatory criteria to be classified as a Professional Client or Market Counterparties and no other Person should act upon it.


Columbia Threadneedle Investments is the global brand name of the Columbia and Threadneedle group of companies.

For professional investors in EMEA APAC only

Legal and regulatory disclosures

The analysis included on this website is for professional investors in EMEA APAC. This material should not be considered as an offer, solicitation, advice or an investment recommendation. This communication is valid at the date of publication and may be subject to change without notice. Any opinions expressed are made as at the date of publication but are subject to change without notice and should not be seen as investment advice. Information obtained from external sources is believed to be reliable but its accuracy or completeness cannot be guaranteed. 

I have read and accept the terms and conditions of this site:

Thank you. You can now visit your preference centre to choose which insights you would like to receive by email.

To view and control which insights you receive from us by email, please visit your preference centre.

Woman listens to music through headphones
Play Video

CT Property Trust- Fund Manager Update

Sed ut perspiciatis unde omnis iste natus error sit voluptatem accusantium doloremque laudantium