April 09, 2026
BCG Predicts Most Jobs Will Be Reshaped, Not Replaced, by AI
One of the most persistent fears surrounding artificial intelligence is that it will hollow out the workforce. New research from Boston Consulting Group (BCG) offers hope to an alternate narrative, concluding that the more immediate reality for most workers is one of transformation rather than elimination. The odds of being reshaped or replaced are not the same for everyone, however. It establishes a hierarchy of job replicability and for roles such as contact centre representative, the outlook is less rosy.
Six Categories of Exposure
BCG’s analysis applies a microeconomic model to approximately 165 million US jobs across 1,500 distinct roles. Its central finding is that between 50 per cent and 55 per cent of jobs will be meaningfully reshaped by AI within the next two to three years. Crucially, this is not the same as being replaced. The vast majority of roles will persist; what changes is the nature of the work itself, the skills required, and the expectations placed on those who perform it.
To map the impact more precisely, the research identifies six categories of exposure. At the optimistic end sit what BCG calls “amplified” roles, where AI augments human capability and demand for the end product expands in parallel. Software engineering is the primary example: AI accelerates code generation and testing, yet system-level judgement and oversight remain firmly human. Around 5 per cent of jobs fall here. A further 14 per cent occupy “rebalanced” roles, where augmentation occurs but demand is more bounded, routine tasks automate, and higher-complexity responsibilities expand. Another 23 per cent sit in ‘enabled’ roles, where AI embeds itself in day-to-day workflows without fundamentally restructuring the job.
The less reassuring categories account for roughly a quarter of all employment. Some 12 per cent of roles are classed as “divergent”, where junior and entry-level positions face the earliest automation of structured tasks while senior responsibilities persist or grow. A further 12 per cent are “substituted”, where AI directly replaces human input in core tasks and demand does not expand to compensate. Across the substituted and divergent categories, BCG estimates that 10 per cent to 15 per cent of all US jobs could be eliminated over the next four to five years.
Substituted: Contact Centre Workers
BCG places contact centre representatives firmly in the substituted category. Most call centre work involves structured, repeatable interactions, such as account lookups, scripted troubleshooting, and policy queries. When agentic AI systems can handle those end to end, fewer representatives are required. At the same time, the overall volume of customer contacts is largely fixed by the size of a company’s customer base, meaning productivity gains are unlikely to generate enough new demand to absorb displaced roles.
This trajectory is evidenced by companies moving to reduce headcount across their contact centre divisions. Sky decided to put 2,000 contact centre jobs at risk, for example, as it moves towards AI-led customer service. More recently, HSBC has been considering cutting 20,000 jobs, predominantly from its contact centres. BCG notes that entry-level roles face the greatest near-term pressure, as the most structured tasks that have historically justified large junior hiring cohorts are precisely those most amenable to automation.
That said, BCG is careful to distinguish substitution from permanent exclusion. As augmented and new roles expand elsewhere in the economy, workers displaced from contact centre positions may transition to those opportunities given appropriate reskilling and mobility support. Some representatives are expected to shift towards higher-value interactions. These might include complex queries, emotionally nuanced escalations, and long-term relationship management that AI systems cannot yet reliably perform. We certainly have not yet seen much of the AI doom prophesised when it comes to jobs. Indeed, a recent in-depth analysis on the topic cites Gartner’s finding that only 11% of service leaders have so far reported frontline reductions.
Balance is the Key
BCG argues that workforce strategy cannot sit downstream of automation planning. It must be embedded directly in competitive strategy. Leaders who focus purely on headcount reduction risk cutting more deeply than AI can realistically substitute, losing institutional knowledge and productivity in the process. Those who fail to rethink work structures, meanwhile, risk ceding ground to competitors who can translate AI productivity gains into faster growth.
The researchers call for deliberate investment in upskilling, reskilling, and structured redeployment pathways, noting that workers will require more frequent capability development as the technology continues to evolve. As research from the World Economic Forum underscores, organisations that act now to reskill their employees will be best positioned for the future.
BCG’s core message is ultimately one of balance. Automation, upskilling, and talent planning must advance together, calibrated to the specific mix of automatable and augmentable work within each business. Getting that balance right, the researchers conclude, will determine both employee prospects and enterprise return on investment.
