AI was expected to help organisations do more with fewer people. But the reality may be far more complicated. Our latest article explores why the “AI will replace jobs” narrative may be incomplete, and what organisations should consider as they navigate AI adoption.
When artificial intelligence (AI), particularly generative AI, began gaining mainstream traction, one of the most common assumptions was that organisations would become dramatically more efficient and therefore require fewer employees. Many executives envisioned leaner operations, lower labour costs, and highly automated workflows that could replace large portions of administrative, operational, and even professional work.
Although some workforce reductions have occurred, the broader reality emerging across industries is more nuanced. Many organisations are not necessarily shrinking because of AI. Instead, they are restructuring, hiring differently, and in several cases expanding headcount to support AI-driven transformation, governance, implementation, and growth. The result is thus an evolving labour market where AI is simultaneously automating tasks, creating new roles, whilst also changing skill requirements, and increasing demand for specialised talent.
The early AI narrative versus the emerging reality
The initial AI narrative was built around productivity gains. AI tools promised to, among other things, automate repetitive work, reduce administrative overhead, streamline decision-making, and enable smaller teams to produce more outputs. Hence, once AI systems matured or could be successfully integrated into organisations, organisations were likely to reduce or consolidate their workforce.
Indeed, several major companies, such as Cisco, IBM, and Salesforce, among others, have publicly discussed workforce changes associated with automation or AI integration, whilst others announced layoffs linked to AI adoption or restructuring around AI initiatives. Companies.
However, as organisations move from experimentation to enterprise-wide AI deployment, many are discovering that AI implementation is labour-intensive in ways they did not initially anticipate. Instead of eliminating the need for people, AI frequently creates demand for additional expertise in areas such as AI integration and deployment, cybersecurity, AI governance and compliance, data engineering, prompt design and workflow engineering, and AI quality assurance, to name a few.
Moreover, recent hiring data suggests that AI-related recruitment has become a major growth driver in many labour markets. According to the Foundit Insights Tracker, AI-linked hiring significantly boosted overall recruitment activity in 2025, with demand projected to continue growing strongly into 2026 (Source: The Economic Times).
Similarly, enterprise hiring studies indicate that organisations adopting AI are increasingly recruiting mid-level professionals with adaptable skills and practical experience rather than simply reducing staff (Source: WGU). In the financial services sector, for example, it has been reported that several major banks have stated that although some functions may shrink, demand for technology, cybersecurity, and AI-related roles continues to grow,
In other words, the relationship between AI and employment is not a straightforward replacement model. The broader organisational reality is that AI systems do not operate independently: they require oversight, integration, maintenance, governance, and continual refinement – which humans can provide.
Risks organisations ought to consider
Despite the opportunities, organisations also face several risks when implementing AI. Four key threats are outlined below.
- Overestimating immediate efficiency gains. Many firms initially underestimate the complexity and effort of integrating AI into real-world workflows. AI deployment often requires substantial investments in infrastructure, governance, security, training, and process redesign before productivity gains fully materialise. Moreover, after the initial enthusiasm with which AI integration would have begun is likely to give way to the tedium, protracted effort and mounting costs, which can often slow down the integration process, and consequently the gains that can be realised.
- Skills gaps. Depending on the organisation, they can find themselves unduly focusing on reducing staff numbers in the areas where AI tools have been introduced. At the same time, organisations are frequently struggling to find employees with sufficient AI expertise. As a result, the growing talent shortages remain a major concern despite fears of workforce displacement (Source: WGU).
- Organisational disruption. AI adoption can create uncertainty, greater resistance, and lower morale among employees concerned about job security, as organisations often do not offer adequate clarity (and assurance as appropriate) about the implications of deeper AI integration.
- Governance and compliance challenges. Finally, cognisant that we are still in the early days of AI use, it must be emphasised that AI systems introduce risks that organisations must be prepared to address in areas including, but not limited to: data privacy; bias and discrimination; intellectual property; cybersecurity; and regulatory compliance.
Considerations for organisations navigating AI in the workplace
The initially envisaged and direct relationship between AI integration and staff reductions, especially in entry-level roles, does not necessarily, nor automatically, appear to produce the anticipated results. For organisations grappling with AI’s potential impact, a balanced and strategic approach is essential.
First, an organisation’s AI strategy should be aligned with its business and/or corporate objectives. Hence, AI initiatives should support broader organisational goals rather than being pursued simply because of market pressure or hype. For example, a critical question that should guide development of the AI strategy is, “What is the business problem we are trying to solve?”
Second, AI rarely replaces entire roles immediately. More commonly, it automates specific tasks within jobs. Hence, organisations should be prepared to analyse workflows at the task level, identify where AI augments human performance, redesign roles thoughtfully, and avoid simplistic assumptions about workforce reductions. Hence, the focus should be on task transformation, rather than just job elimination.
Third, AI adoption should be paired with deliberate employee development initiatives. Priority areas may include AI literacy training, digital skills development, data analysis capabilities, critical thinking and problem-solving, and change management readiness. Ultimately, by investing in workforce reskilling, which is often less costly and less disruptive than large-scale replacement hiring, organisations can also improve staff morale and their corporate culture.
Fourth, AI implementation should not be treated solely as an IT project, which is often a fatal mistake organisations make. The approach should be one of cross-functional AI governance that involves a wide cross-section of teams, including the human resource department, legal and compliance, cybersecurity, operations, communications, and executive leadership. Implementing cross-functional governance allows the organisation to draw from a broad range of expertise and functional areas, thus allowing it to better manage both operational and ethical risks.
Finally, organisations should resist the temptation to over-automate critical processes. Noting some of the weaknesses of current AI models, especially regarding bias and data privacy, human oversight is vital, particularly in areas involving hiring, performance evaluations, customer relationships, financial decisions, healthcare, and sensitive public-facing communications.
Final thoughts
The widespread assumption that AI would simply reduce workforce size is proving incomplete. Although AI is undoubtedly automating certain functions and contributing to workforce restructuring, it is also generating substantial new hiring demand, particularly for specialised, hybrid, and governance-related roles. In many organisations, AI is not replacing people outright. It is changing the nature of work, altering skill requirements, and expanding the need for human oversight, integration, and strategic management.
The organisations most likely to succeed will not necessarily be those that replace the most workers with AI, but those that most effectively combine human capability with AI-driven tools and processes. Ultimately, the future of work may depend less on ‘humans versus AI’ and more on how successfully organisations develop and nurture ‘humans working with AI’.
Image credit: Gerd Altmann (Pixabay)
