Will your job still exist in 2040? AI and the future of employment

In summary

• Job loss fears persist as AI continues to transform workplaces, but evidence suggests AI is more likely to see jobs evolve than be eliminated 
• It presents significant challenges for organisations around training, scaling and governance for AI 
• Crucially, organisations must manage massive uncertainty over the technology’s future capabilities, capacity and costs 
• Flexible and dynamic workforce management, processes and AI strategies will be essential to get the best from people and the changing technology 

People aren’t hopeful. A new survey by King’s College London in May found that a majority in the UK think AI will hurt more than it helps in the jobs market. Seven in ten are worried by the economic impacts of the technology, the university found, with six in ten saying it will create fewer jobs than it destroys. One in five think it will lead to civil unrest.  

There’s already some evidence to support at least the first two propositions. Earlier in the month, Challenger, Gray & Christmas, a provider of “outplacement” services supporting employees who have been laid off, said AI was the leading cause of job cuts in April for the second consecutive month. It accounted for more than a quarter of the total.  By 2040, some analysis suggests, US unemployment could double to 8%. 

The UK, meanwhile, with its heavily service-based economy, may be more exposed still. A report by City Hall in April suggests that at least a million jobs in London are significantly vulnerable to the impact of AI.   

There are, however, caveats to that. For a start, it’s clear that impacts won’t be evenly felt.  

There is significant variation in both the breadth and depth of AI adoption across industries. Microsoft’s analysis shows that sectors such as retail or education, and particularly software and technology, lead in the share of organisations using AI agents; in others, however, such as manufacturing and resources, AI agents are less common but, where present, are deployed at far greater scale.  

“The real difference lies in where agents are embedded and how extensively organisations have integrated them into their workflow,” the technology giant suggests.  

Job predictions can be overblown. Regardless of whether AI is the principal cause of US layoffs, for instance, there’s little sign of an impending apocalypse. The unemployment rate in the country remained steady in April. The UK government’s assessment notes that several analyses show job postings declining more sharply in occupations with higher AI exposures, but declines are concentrated in junior jobs in high-wage sectors; technical roles, such as software engineers and data analysts, seem particularly at risk.  

It also notes that there’s little certainty as to the real impact. “Other evidence finds no significant employment effects from AI adoption to date and highlights alternative explanations,” it remarks. 

Certainly, the pessimism ignores the benefits AI may bring. A World Economic Forum’s Future of Jobs report last year found that 86% of respondents expected AI to transform their business by 2030. Together with trends in demographics and the green transition, technological change (led by AI and automation) would displace 92 million jobs by 2030, it predicted. But it would generate 170 million new roles, too. 

Moreover, that may overstate the pace of change. Historically, radical change from new technologies, such as steam and electricity haven’t happened overnight. Rather than rapid job losses, the more compelling case for now is that AI is redefining roles:  

• Changing how workers do their jobs  
• Boosting productivity  
• Expanding the range of employees who can do high-value work  
• Altering the skills employees need 

Crucially, the research suggests the key challenge AI poses is as much to employers as employees. As Microsoft puts it, “[A] growing share of workers are using AI in advanced, resourceful ways. The problem? Most organisations aren’t keeping up. 

“In many cases, people are ready. The systems around them are not.” 

For that to change, businesses must address several significant challenges.  

Perhaps the most obvious need is for new skills. Workers do not necessarily need specialist AI skills but rather a broader toolkit to enable them to adapt to the new ways of working and the expanded possibilities for productivity and higher-value work the technology brings.  

“Skills that require collaboration with colleagues, originality, and experience with basic office tools, have seen the most significant increase in demand in occupations highly exposed to AI,” the OECD notes. 

Its research found that 72% of vacancies in occupations with high AI exposure required at least one management skill, and more than two-thirds (67%) required at least one skill related to business processes. Social, emotional and digital skills were also in high demand. As the saying goes – AI won’t take your job, but someone who uses it might. 

Even if staff development isn’t recognised as an obvious organisational imperative to secure the benefits AI can bring in the workplace, it’s increasingly a prerequisite for recruitment and retention. Surveys show consistently high demand for training from staff to enable them to adapt to AI.   

On the other hand, it will be difficult to effectively hire or train staff without some understanding of how AI is transforming their roles to begin with. At present, while the use of AI tools can be widespread in businesses, it is frequently unstructured and, in many cases, invisible to the wider organisation. Much AI use remains informal, with individuals simply harnessing consumer AI tools or features embedded in existing software platforms to ease and enhance their work.  

This is bringing valuable (if limited) productivity and efficiency gains to organisations, but not transformative change. There’s often little connection between the high-level corporate AI strategy and vision, and the practice on the ground.  

Even where use is visible and sanctioned, successful pilots or small-scale adoptions are often mistaken for evidence of organisational adoption. This ignores the informal effort and crucial role of individuals in many of these projects, and how difficult they can be to replicate across large organisations with legacy systems, controls and ways of working.  

This is a similar point to that identified by the UK Department for Science, Innovation and Technology (DSIT) when considering evidence from occupation-level studies.  

“The emerging evidence from these studies is a useful indicator of potential but is not a guarantee of large-scale productivity gains. Experimental settings differ substantially from real-world workplaces, tasks are often more clearly specified, and environments more controlled,” it noted.  

“There is also likely to be a period of organisational adjustment in which firms experiment with AI tools, integrate them into workflows, and train staff, which may temporarily slow measured productivity,” it adds. 

Closely related to these discussions is the issue of AI governance. The lack of visibility of AI use across organisations doesn’t just limit opportunities to capture and scale innovation and good practice; it also prevents proper control to confidently manage the risks.  

Appropriate guardrails are not about restricting innovation but rather defining risk tolerances and providing a framework within which AI use can safely flourish. As Colin Bell, Chief Compliance Officer for HSBC Group, explained at an S&W event earlier this year, effective governance enables those within an organisation to experiment with the technology safely and confidently.  

 “Often, you’re looking to loosen the organisation up to adopt technology at pace,” he said.

To do that, however, a business must understand where and how AI is being used within the organisation. They also need to know how the AI systems act. AI observability – the ability to monitor, understand and explain how AI systems behave in real time – is the foundation for effective governance. And in many organisations, it is badly underdeveloped. 

But if organisations have poor visibility of current AI use, that’s more than they have of its future. The final challenge for businesses is the massive uncertainty that remains. Adoption of AI across most businesses remains in a nascent stage. The technology itself is also still young, developing rapidly and facing a potentially volatile future given fears of a bubble in AI investment.  

Huge questions remain over both capabilities and capacity. With regard to the former, we are still some way from understanding the technology’s limitations and the extent to which these may be overcome in time.  

As the DSIT report put it: “There is uncertainty over whether AI agents can be trained to reliably complete complex tasks across a broad range of domains. If this were to occur, it could have significant implications for the labour market, though the nature and scale of such impacts would depend on many factors, including adoption rates, workforce adaptation and policy responses.” 

AI capacity, meanwhile, is limited by hardware and energy constraints, either absolutely or through pricing: Witness Anthropic restricting access to Claude at peak times. The content AI relies on to train its models is also increasingly gated, creating another challenge for the industry. It is futile to try to predict how the AI market and technology will evolve over the coming decade or more, but we should acknowledge that the AI offerings today may be very different in years to come. This will be a significant factor in shaping the workforce given that costs, as much as capabilities, determine the extent to which automation replaces human labour.  

All this will have a profound effect on AI’s applications and impact.  

Consequently, the challenge to organisations is as profound as it is for individual employees. AI is changing the way successful businesses must work:  

• They will need to invest in skills and training to enable workers to use AI tools effectively, and to develop the wider skills they may need as AI changes their roles 
• They must improve visibility of AI use, and learn how to identify and scale successful uses in pockets of the workforce to roll them out enterprise-wide   
• They must put in place effective governance and risk frameworks to ensure appropriate controls and compliance, while enabling and empowering innovation  

Crucially, they will also need more dynamic, flexible and adaptable workforce, operations, data, technology and digital transformation strategies. Faced with an uncertain future, they must be able to respond effectively as evidence emerges of what AI can do, what it can’t, where it can add value, and how its capabilities, capacity and costs are evolving. And they’ll have to do so continually. 

However the workforce changes, for businesses trying to get the best from their people and technology in the age of AI, the job is never done.