Artificial intelligence is transforming the modern workplace. By automating various organizational processes, A.I. challenges and even replaces certain roles, significantly impacting the technology sector. According to estimates from Boston Consulting Group (BCG), artificial intelligence will reshape 50% to 55% of US jobs within just the next 3 years. This estimate comes from an analysis of tasks associated with 1,500 distinct occupations to evaluate which specific roles are most compatible with AI augmentation or are at risk of total replacement.
BCG researchers emphasize that task automation does not always translate to total job losses. Rather, they have assessed the future of work using 3 particular lenses: task-level automation potential, the dynamics between substitution and augmentation, and the degree of demand expandability within a given field. Although the study notes that some job roles will inevitably become obsolete, it also emphasizes how fundamentally most positions will change in their function. BCG projects that approximately 10% to 15% of jobs in the U.S. could be replaced by A.I. within the next 5 years.
Matthew Kropp, BCG Managing Director and Senior Partner, told CBS News that even if roles continue, the tasks involved will transform. He warned against impulsive, widespread layoffs, calling them a “knee-jerk reaction” harmful to both the workforce and organizations. Instead, Kropp emphasized that the future will require companies to reskill employees, urging corporate leaders to focus on augmenting human labor with AI instead of using the technology to replace workers.
What the BCG Analysis Actually Found
The Boston Consulting Group (BCG) categorizes these shifts into specific roles: amplified, rebalanced, and enabled. In these categories, AI enhances productivity and shifts focus away from monotonous, repetitive tasks. BCG analyzes jobs based on the nature of the change. “Amplified” positions increase output while maintaining their core objectives. “Rebalanced” roles adjust their task structure to reduce routine labor. “Enabled” roles experience significant daily changes due to high automation potential, even without changes in staffing levels. While each category undergoes substantial transformation, the key takeaway is that the roles themselves do not necessarily vanish.
For instance, demand for “amplified” roles like software engineering may rise as falling costs increase the need for experienced engineers to manage AI-generated code and concentrate on architecture. However, in contrast, call centers are expected to become obsolete as AI expansion and usage do not expand demand. “When AI reduces the cost of routine inquiries, interaction volume does not grow proportionally. In this context, productivity gains are more likely to reduce the number of representatives required,” which BCG disclosed in their analysis. Ultimately, BCG projects that while most positions will change fundamentally in function, only approximately 10% to 15% of US jobs may be replaced by AI within the next 5 years.
Goldman Sachs Adds Global Scale
According to Goldman Sachs Research, artificial intelligence has the potential to replace 300 million full-time positions on a global scale. Within the United States and Europe, approximately 25% of all work hours are subject to being fully automated. Economist Joseph Briggs from Goldman Sachs anticipates that between 6% and 7% of the American workforce could experience displacement, which may lead to a 0.6 percentage point increase in unemployment over the next decade. While a faster rate of AI adoption could lead to more significant changes in the employment landscape, Goldman Sachs also highlights that this technology could concurrently spark a substantial surge in productivity.
The Net Numbers From the WEF and IMF
The IMF estimates roughly 40% of global jobs carry some AI exposure. In advanced economies, that figure climbs to 60%. The World Economic Forum projects AI will displace 92 million roles by 2030 while creating 170 million new ones. That leaves a net gain of 78 million positions globally. However, the people displaced and those who gain new roles are often not the same workers. The transition creates winners and losers simultaneously, and the gap between them is widening fast.
The Jobs Going and the Jobs Growing
BCG identifies call centers as a primary example of how AI efficiency does not necessarily lead to increased demand. According to Gartner, companies are expected to save $80 billion in call center labor expenses by 2026 due to AI-driven automation. Furthermore, research from SupportYourApp suggests that by the end of 2026, generative AI will manage 1 out of every 10 customer support interactions. Consequently, entry-level customer service representatives are among the most vulnerable to these technological shifts.
Software Engineering Moves in the Opposite Direction
Software engineering trends in the opposite direction: as the cost of jobs decreases, the demand for these positions increases. Matthew Kropp, BCG Managing Director and Senior Partner, emphasized this shift as a significant change in the tech sector. The hiring platform TrueUp reports that over 67,000 software engineering positions will be vacant at the beginning of 2026, reflecting an approximate 30% increase since the start of the year. As AI lowers the cost of routine implementation, organizations shift their focus; they increasingly prioritize experienced engineers to manage AI-generated code and tackle complex challenges.
What the University of Washington Told Its Students
According to a CNN report, Magdalena Balazinska, who serves as the director of the Paul G. Allen School, addressed a group of more than 2,000 undergraduates regarding the impact of AI on employment. She emphasized that rather than reducing the number of available jobs, AI is actually expanding professional opportunities. This sentiment was echoed by CoderPad CEO Amanda Richardson, who informed CNN that while the fundamental nature of work is evolving, the roles themselves are not being eliminated. Richardson further noted that leading engineers are already integrating AI into their daily workflows to refine their designs and elevate the overall quality of their contributions.
The Roles AI Cannot Reach
Certain professions remain shielded from the reach of artificial intelligence, particularly those that require a physical presence or nuanced interpersonal evaluation. Matthew Kropp highlighted therapists and plumbers as examples of roles unlikely to face major disruption from AI. At the same time, the global job market sees a surge in AI-centric opportunities, with LinkedIn reporting the addition of 1.3 million new positions, including data annotators, AI engineers, and forward-deployed engineers, within a two-year span. By 2026, Mashable recognized AI engineer as the fastest-growing technical role on the LinkedIn platform. Looking toward the future, the World Economic Forum anticipates that AI integration will create 170 million new jobs across various industries by 2030.
The Reskilling Crisis Companies Are Ignoring
Matthew Kropp issued a blunt caution to CBS News, stating that businesses often respond to AI by implementing broad, non-selective job cuts. He argued that this tactic harms both the workforce and the organizations. Research from Gartner supports this view, revealing that only 20% of enterprise clients have reduced their staff specifically due to AI recently. In early 2026, many job losses attributed to AI were actually driven by broader economic factors, highlighting a crucial distinction since these 2 scenarios require vastly different strategic approaches.
Forbes reported that companies building reskilling academies around cloud, data, AI, and cybersecurity saw 22% to 30% higher internal mobility. They also cut external hiring costs by 15% to 20%. The World Economic Forum reports 85% of employers plan to prioritize workforce upskilling by 2030. Gartner adds that 80% of engineers alone will need to upskill through 2027. That is the pace of generative AI’s evolution. Companies treating learning as a core function are building workforces that hold up.
The IMF Data on Skill Adoption and Employment
The IMF found that AI skill adoption raised regional employment by 1.3%. That rise occurred per one-point increase in skill-related job postings over a decade. Workers combining AI fluency with human capabilities earn measurably more than peers without those skills. LinkedIn data shows a 70% year-over-year increase in US roles requiring AI literacy. That demand is not slowing. Companies absorbing this shift early carry a structural hiring advantage over those who wait.
The Scale of What Needs to Happen
The World Economic Forum estimates 80% of the global workforce will need new skills by 2027 to remain competitive. An estimated 120 million workers face medium-term redundancy because reskilling will not reach them in time. McKinsey found 40% of roles now require major skill upgrades. Companies that reskill well are 2.5 times more likely to retain critical talent. Kropp said the priority must be moving people into areas where jobs remain stable, not eliminating their positions entirely.
Who Gets Hit First, and What Comes Next
Entry-level workers carry the sharpest immediate exposure. BCG’s analysis identified junior positions in divergent roles as the first to be automated. Structured tasks traditionally performed at the entry level are among the earliest AI can absorb. Goldman Sachs data shows workers aged 22 to 25 in AI-exposed roles have seen a 16% employment drop. Experienced workers in the same fields remain stable. Cornell University research found US companies adopting AI reduced junior hiring by approximately 13%.
The Pipeline Problem Nobody Is Talking About
The traditional on-ramp into knowledge industries is narrowing. Junior roles that train workers through structured repetitive tasks are being automated before those workers gain experience to move up. This creates a pipeline problem, not just an employment one. Dario Amodei, CEO of Anthropic, warned that AI could eliminate 50% of entry-level white-collar jobs within a few years. The New York Times reported that economists who once dismissed AI job threats are no longer doing so.
New Jobs That Did Not Exist Before
However, AI is generating entirely new categories of work. LinkedIn data shows AI engineer became one of the fastest-growing jobs on the platform over 3 years. The global economy added 1.3 million new AI-related roles in two years. More than 600,000 of those were data center positions. Roles like AI trainer, prompt engineer, and human-AI collaboration specialist did not exist at scale five years ago. They are now among the most actively recruited positions in technology globally.
The Historical Pattern
Every major technological transition triggered similar predictions of mass unemployment. The printing press, the industrial revolution, electrification, and the internet all generated credible warnings about job destruction. Each time, the technology reshaped work rather than eliminating it completely. New industries emerged and new jobs formed around the technology itself. Workers who adapted to the transition captured the gains as opposed to workers who failed to adapt and change over, faced prolonged hardship.
Kropp uses social media as his reference point. When platforms like Facebook and Instagram launched, nobody anticipated that social media influencers would become a legitimate full-time career. The role did not exist. The technology created the conditions for its existence. AI will likely generate analogous job categories that currently have no name or established precedent. Predicting what those jobs will look like is genuinely difficult at this stage.
The Bureau of Labor Statistics incorporated AI impacts into its official employment projections for the first time in 2025. The projections reflect sectoral growth in AI-adjacent fields alongside contraction in high-automation-exposure categories. This marks the first time a U.S. government agency embedded AI as a structural variable in official employment forecasts, signaling how seriously federal labor economists now treat the scale of disruption ahead.
Read More: New AI Tool Might Predict Pancreatic Cancer Risk Years Before Diagnosis
What Employers Must Do Now
BCG’s position on employer strategy is direct. Kropp told CBS News the focus must be reskilling, not replacement. He said companies should move people into areas where jobs remain stable. IBM and IDC research projects that companies could lose $5.5 trillion by 2026 due to skills gaps in AI transition planning. McKinsey found that firms investing in human-AI collaboration produce better outcomes than those using AI primarily to reduce headcount. The margin between those two strategies is growing each quarter.
The Cost of Getting It Wrong
Employers feeling the consequences of rushed cuts have documented the cost. Gartner projects half of all companies that attributed staff reductions to AI will rehire for similar roles by 2027. Forrester found 55% of employers already regret AI-related workforce reductions. Rehiring costs more than retention does. Companies that treated AI adoption as a cost-cutting exercise lost institutional knowledge alongside headcount. They are now competing to rebuild what they voluntarily dismantled, often in a tighter talent market.
How the Best Companies Are Building for This
The most effective reskilling strategies share a common structure. Forbes Business Council members identified transparency as the first requirement. Companies must communicate clearly that change is coming and that the goal is growth. The second is embedding training into daily workflows rather than treating it as a separate HR program. BrightStar automated insurance credentialing and payroll, then redirected staff toward higher-value client work. Employees stayed engaged because the work became more meaningful, not because job security was simply promised.
The Timeline Is Already Running
Digital Applied research shows a structured upskilling plan can move a worker up one AI skill tier in 60 days. Skills in AI-exposed roles evolve 66% faster than in other jobs. LinkedIn data shows 53% of US employees plan to learn new AI skills within six months. A further 48% believe those skills will help them advance. The workforce is already moving in this direction. Companies not moving with it are falling behind in real time
A.I. Disclaimer: This article was created with AI assistance and edited by a human for accuracy and clarity.
Read More: 9 Myths About Intelligence You Probably Still Believe