Skip to main content Skip to footer
  • "com.cts.aem.core.models.NavigationItem@5fe2b26b" Careers
  • "com.cts.aem.core.models.NavigationItem@10f83f3b" News
  • "com.cts.aem.core.models.NavigationItem@5f136db4" Events
  • "com.cts.aem.core.models.NavigationItem@3fd25fdb" Investors

New work, new world 2026:

How AI is reshaping work faster than expected

Our freshly updated research on AI and jobs reveals disruption that is more extensive—and swift—than we anticipated three years ago. What we predicted would happen in a decade’s time is already here.
Download PDF

Executive summary

A few years ago, we made news with a study predicting an astounding 90% of jobs would be disrupted by AI in less than a decade’s time. As it turns out, though, we underestimated the impact of the technology.

What we projected might take until 2032 to unfold is happening now before our eyes. Today—six years ahead of schedule—93% of jobs could be impacted in some way by AI. In the US alone, this could add up to about $4.5 trillion worth of labor shifting from humans to AI. The technology, in short, is affecting more jobs, faster, and to a greater extent than we anticipated.

We came to these findings by updating our 2023 research on AI and jobs, in which we assessed 18,000 tasks performed by 1,000 professions in terms of the degree to which they could be automated or assisted by AI. A lot has happened in the ensuing three years.

Since then, AI models have become increasingly adept at interpreting many types of input, including images, diagrams and video. Further, more sophisticated AI models have emerged with advanced reasoning capabilities. And finally, AI agent-driven systems are now capable of completing complex workflows with minimal human oversight.

With these three advancements—multimodality, advanced reasoning and agentic AI—it was time for a new look at how AI could reshape the workforce. So, we conducted a thorough reevaluation of the 18,000 tasks, this time through the lens of AI’s enhanced potential to assist or automate them.  

What we found: Across all occupations, average exposure scores (i.e., the degree to which an occupation could be affected by AI) are an astounding 30% higher than what we’d forecast they’d be by 2032. (See explainer box for more on the exposure score.)

In fact, while our original analysis found an average 2% annual increase in exposure scores among the jobs studied, we are now seeing a 9% annual score increase. As a result, some jobs that seemed safe from change when large language models (LLMs) first became mainstream are now capable of being affected much more quickly (see Figure 1). 

Understanding the $4.5 trillion economic impact of AI

We wanted to quantify the economic value of the total amount of work that AI could assist with or automate today. To do that, we used data from the US Bureau of Labor Statistics on the number of employees in each of the occupation groups in our study and then multiplied that by these employees’ median annual salaries. Then, using our exposure scores, we assessed the amount of this total economic value that could theoretically be exposed to AI.

The result—$4.5 trillion—is based on the assumption that the traditional way a task is completed would shift seamlessly to AI. Despite being theoretical, however, the calculation provides a glimpse into the sweeping economic change AI could bring.

A few years ago, we made news with a study predicting an astounding 90% of jobs would be disrupted by AI in less than a decade’s time. As it turns out, though, we underestimated the impact of the technology.

What we projected might take until 2032 to unfold is happening now before our eyes. Today—six years ahead of schedule—93% of jobs could be impacted in some way by AI. In the US alone, this could add up to about $4.5 trillion worth of labor shifting from humans to AI. The technology, in short, is affecting more jobs, faster, and to a greater extent than we anticipated.

Understanding the $4.5 trillion economic impact of AI

We wanted to quantify the economic value of the total amount of work that AI could assist with or automate today. To do that, we used data from the US Bureau of Labor Statistics on the number of employees in each of the occupation groups in our study and then multiplied that by these employees’ median annual salaries. Then, using our exposure scores, we assessed the amount of this total economic value that could theoretically be exposed to AI.

The result—$4.5 trillion—is based on the assumption that the traditional way a task is completed would shift seamlessly to AI. Despite being theoretical, however, the calculation provides a glimpse into the sweeping economic change AI could bring.

We came to these findings by updating our 2023 research on AI and jobs, in which we assessed 18,000 tasks performed by 1,000 professions in terms of the degree to which they could be automated or assisted by AI. A lot has happened in the ensuing three years.

Since then, AI models have become increasingly adept at interpreting many types of input, including images, diagrams and video. Further, more sophisticated AI models have emerged with advanced reasoning capabilities. And finally, AI agent-driven systems are now capable of completing complex workflows with minimal human oversight.

With these three advancements—multimodality, advanced reasoning and agentic AI—it was time for a new look at how AI could reshape the workforce. So, we conducted a thorough reevaluation of the 18,000 tasks, this time through the lens of AI’s enhanced potential to assist or automate them.  

What we found: Across all occupations, average exposure scores (i.e., the degree to which an occupation could be affected by AI) are an astounding 30% higher than what we’d forecast they’d be by 2032. (See explainer box for more on the exposure score.)

In fact, while our original analysis found an average 2% annual increase in exposure scores among the jobs studied, we are now seeing a 9% annual score increase. As a result, some jobs that seemed safe from change when large language models (LLMs) first became mainstream are now capable of being affected much more quickly (see Figure 1). 

Our 10-year forecast is happening today

Across the board, AI-driven change is both more extensive and happening more quickly than anticipated.

Our 10-year forecast is happening today

Figure 1 

Source: Cognizant

Four consumer groups, based on AI intensity

Figure 1 

Source: Cognizant

Understanding exposure scores

To calculate the exposure score, we used the same approach as in our original research. We examined 18,000 tasks and close to 1,000 jobs in the O*NET database, assessing the tasks for automatability on a five-point scale (not automatable, minimally AI-assistable, partially AI-assistable, mostly AI-assistable and fully automatable).

In addition to how many of the job’s tasks could be automated or assisted by AI, we also considered the relative importance of the task.

We used an AI model to get an initial assessment of task classification but then reviewed and reclassified the findings where necessary.

Using that analysis, we calculated an exposure score for each profession. The score reflects the degree to which an occupation could be affected by AI. A higher score means a higher percentage of the job's tasks are assistable or automatable, so people in that profession could be greatly affected.

The analysis in this report is based on a fresh assessment of AI’s capabilities, reflecting the rapid evolution of the technology over the past three years, specifically its multimodal, reasoning and agentic capabilities.

The resulting exposure scores represent a theoretical maximum: what current AI technology could potentially accomplish with optimal implementation. The scores do not account for enterprise adoption, employee acceptance, regulatory frameworks, quality control requirements, ethical considerations or the substantial organizational change required to deploy AI at scale.

For these reasons, the exposure score represents a raw calculation of the technology’s potential. As such, it reflects capability and opportunity rather than inevitability. Throughout the report, the term “exposure score” is a theoretical exposure score. 

More jobs are more highly exposed

The percent of jobs with the lowest exposure has shrunk from 31% to 7%, while the percent with the highest exposure has grown from 0% to 30%.

More jobs are more highly exposed

Figure 2
Source: Cognizant

To understand which jobs and job families are experiencing the fastest surge in AI exposure, we also calculated a velocity score, which quantifies the difference between the original trajectory of change in exposure scores over time and the new trajectory based on our refreshed analysis. (See explainer box for more on the velocity score.)

Examples of occupations showing unexpectedly high velocity scores, particularly when compared with their exposure levels in the original research, include: 

  • Roles dominated by manual-labor tasks. Once considered a safe haven from AI disruption, many jobs requiring a great degree of physical labor now show significantly higher exposure scores compared with our original research, along with unexpectedly fast velocity.

    In construction, for example, AI can now help with interpreting blueprints. In transportation, it can inspect shipments or conduct safety reviews. The idea that a car mechanic or plumber can put on a pair of AI-augmented glasses to assist in locating a faulty engine part or leaking pipe is now far from science fiction.  

  • Decision-making roles. Managerial and supervisor jobs are now increasingly exposed due to the emergence of agentic AI. Previously, these roles were more insulated from disruption because they involve complex coordination and judgment. Agentic AI alters this dynamic by moving beyond analysis to execution.

    Where managers once spent significant time allocating resources, monitoring project status or triaging workflows, autonomous agents can now orchestrate these duties. Project managers, for example, could rely on agents to autonomously schedule meetings, reallocate budget based on spend patterns and chase status updates by leveraging tools they are integrated with.

  • Hyper-specialized sectors, such as healthcare, education and law. In these sectors, AI has quickly moved from assisting with low-level tasks to automating more complex tasks that are critical to the role.

    For instance, AI is revolutionizing healthcare by improving diagnostic accuracy and supporting patient care. In education, it can facilitate student assessment and classroom discussions. In law, it can analyze probable outcomes and assist with contract negotiations.
About the velocity score

To identify the roles and occupation groups that have seen the most accelerated forms of disruption, we developed an additional measurement: the velocity score. The velocity score represents the difference between the original annual acceleration rate of exposure scores for any given job and the updated rate.

This metric reflects how fast the pace of change could be for any given occupation, given the most recent advances in AI.

A low score indicates that the latest changes in AI have impacted the role to a relatively minor extent. A high score reveals that the latest innovations will impact the role significantly.

About the velocity score

In this report, we identify the biggest AI advancements in the last three years and why this has accelerated job impact. We also highlight the job families that could see the most—and the fastest—change, as well as some job groups in which change may be less dramatic but is still more extensive than originally anticipated.

We also provide guidance on how business leaders can navigate the changes ahead. By embracing the following mindset shifts, businesses can better plan for the disruption to their workforce that is happening more quickly than imagined.

Three short years; three big changes in AI capabilities

The pace of change in work is now inextricably tied to the acceleration of AI itself. In 2023, most LLMs in use by businesses operated like narrow savants. They could generate text and code with fluency but had little grasp of planning, context or consequence.

Today, advances in AI capabilities enable entirely new types of work to be automated or assisted by AI. 

Consider that in our analysis, one-third of all occupational tasks remain “not automatable.” However, the percent of tasks we classified as “fully automatable” has risen to 10% from 1% three years ago—and just two percentage points short of the 12% originally forecast for 2032.

Even more revealing, nearly 40% of all tasks can now be classified as being “partially” or “mostly” assistable by AI vs. just 15% previously. This also exceeds the 2032 forecast of 31%.  This middle category is where change is most intense, as agentic systems become operational (see Figure 3). 

Percent of tasks classified as “fully automatable”

1%

Original

10%

Today

Percent of tasks that are partially or mostly assistable

15%

Original

40%

Today

More tasks are more automatable by AI
More tasks are more automatable by AI

Figure 3 

Source: Cognizant

With that in mind, here are the three key AI capabilities we considered when updating our job exposure scores:

1. Multimodal AI: Creating systems that see

Multimodal models provide AI with the eyes and ears that help connect digital systems with the physical world. These models can parse images, diagrams and video, recognize spatial relationships, and cross-reference visual data with text or numerical inputs. Where traditional AI could only describe the world, multimodal AI can interpret it.

This new digital-physical connection has real occupational consequences. Jobs involving design review, product testing and quality control were previously beyond AI’s reach because they relied on visual comprehension.

Now, true multimodal models can evaluate design layouts, identify defects in manufacturing lines and assess the completeness of building construction from site photographs. Combined with sensor data and robotic integration, multimodality extends automation into the tactile and perceptual fabric of work. As a result, these types of jobs have climbed the exposure scale sharply.

AI can now understand images, diagrams, video and spatial relationships

AI can now understand images, diagrams, video and spatial relationships

Jobs involving design review, product testing, maintenance and quality control are now more highly exposed

Jobs involving design review, product testing, maintenance and quality control are now more highly exposed

AI can now understand images, diagrams, video and spatial relationships

AI can now understand images, diagrams, video and spatial relationships

Jobs involving design review, product testing, maintenance and quality control are now more highly exposed

Jobs involving design review, product testing, maintenance and quality control are now more highly exposed

Multimodal models provide AI with the eyes and ears that help connect digital systems with the physical world. These models can parse images, diagrams and video, recognize spatial relationships, and cross-reference visual data with text or numerical inputs. Where traditional AI could only describe the world, multimodal AI can interpret it.

This new digital-physical connection has real occupational consequences. Jobs involving design review, product testing and quality control were previously beyond AI’s reach because they relied on visual comprehension.

Now, true multimodal models can evaluate design layouts, identify defects in manufacturing lines and assess the completeness of building construction from site photographs. Combined with sensor data and robotic integration, multimodality extends automation into the tactile and perceptual fabric of work. As a result, these types of jobs have climbed the exposure scale sharply.

2. Expanded AI reasoning: Creating systems that think

Reasoning was once the missing ingredient in AI’s cognitive repertoire. Early generative models produced fluent language but faltered on multistep logic or long-term coherence.

The breakthrough came with structured reasoning frameworks and reinforcement-style fine-tuning. This culminated in models that demonstrate consistent, transparent chains of thought, enabling them to test hypotheses, deconstruct problems and evaluate alternative strategies.

This reasoning ability has reclassified entire clusters of cognitive work. Analytic tasks, such as those found in consulting, finance and law, have shifted from being partially to mostly assistable by AI. For instance, a market analyst can now prompt an AI to not only summarize market data but also identify outliers, construct scenario models and justify recommendations with evidence. Audit and compliance tasks could now be entirely executed by reasoning agents that understand both numerical logic and procedural context. Planning, forecasting and diagnostic problem-solving are now within the operational domain of AI systems.

New reasoning models can tackle complex cognitive activities

New reasoning models can tackle complex cognitive activities

Exposure levels have risen sharply for people who do planning, forecasting and diagnostic problem-solving

Exposure levels have risen sharply for people who do planning, forecasting and diagnostic problem-solving

New reasoning models can tackle complex cognitive activities

New reasoning models can tackle complex cognitive activities

Exposure levels have risen sharply for people who do planning, forecasting and diagnostic problem-solving

Exposure levels have risen sharply for people who do planning, forecasting and diagnostic problem-solving

Reasoning was once the missing ingredient in AI’s cognitive repertoire. Early generative models produced fluent language but faltered on multistep logic or long-term coherence.

The breakthrough came with structured reasoning frameworks and reinforcement-style fine-tuning. This culminated in models that demonstrate consistent, transparent chains of thought, enabling them to test hypotheses, deconstruct problems and evaluate alternative strategies.

This reasoning ability has reclassified entire clusters of cognitive work. Analytic tasks, such as those found in consulting, finance and law, have shifted from being partially to mostly assistable by AI. For instance, a market analyst can now prompt an AI to not only summarize market data but also identify outliers, construct scenario models and justify recommendations with evidence. Audit and compliance tasks could now be entirely executed by reasoning agents that understand both numerical logic and procedural context. Planning, forecasting and diagnostic problem-solving are now within the operational domain of AI systems.

3. Agentic AI: Creating systems that act

The defining feature of AI in the post-2024 landscape is its agentic capability. If multimodality gives AI eyes and ears, and reasoning expands AI’s mental map, agentic capabilities give it hands. This is because AI systems no longer stop at generation; they can take meaningful action.

Consider that new technologies such as Model Context Protocol servers, intelligent function-calling systems and secure tool integration now allow AI to work directly with core business platforms. AI agents can now work together to fetch live data, execute commands in third-party software and monitor results for feedback loops.

For example, multiple marketing agents can plan a campaign, query databases for segmentation, create advertising assets, schedule social media posts and report performance, all through connected tools.

This agentic capability has pulled many administrative and coordination tasks further into the high exposure score zone. Schedulers, office administrators and project assistants once saw limited exposure because AI could not manipulate enterprise software directly. Now, the boundary between “knowledge work” and “process work” is fading as systems handle execution as well as instruction.

Agentic capabilities have also changed management itself. Supervisory tasks, such as allocating work, checking progress and escalating issues, can increasingly be mediated by autonomous systems. In hybrid environments, teams of human and machine agents already collaborate across shared dashboards, with AI handling workflow triage and exception management.

Supported by new technologies, AI systems can now take meaningful action

Supported by new technologies, AI systems can now take meaningful action

Schedulers, office administrators, project assistants and supervisory roles have moved from limited to high exposure levels

Exposure levels have risen sharply for people who do planning, forecasting and diagnostic problem-solving

Supported by new technologies, AI systems can now take meaningful action

Supported by new technologies, AI systems can now take meaningful action

Schedulers, office administrators, project assistants and supervisory roles have moved from limited to high exposure levels

Exposure levels have risen sharply for people who do planning, forecasting and diagnostic problem-solving

The defining feature of AI in the post-2024 landscape is its agentic capability. If multimodality gives AI eyes and ears, and reasoning expands AI’s mental map, agentic capabilities give it hands. This is because AI systems no longer stop at generation; they can take meaningful action.

Consider that new technologies such as Model Context Protocol servers, intelligent function-calling systems and secure tool integration now allow AI to work directly with core business platforms. AI agents can now work together to fetch live data, execute commands in third-party software and monitor results for feedback loops.

For example, multiple marketing agents can plan a campaign, query databases for segmentation, create advertising assets, schedule social media posts and report performance, all through connected tools.

This agentic capability has pulled many administrative and coordination tasks further into the high exposure score zone. Schedulers, office administrators and project assistants once saw limited exposure because AI could not manipulate enterprise software directly. Now, the boundary between “knowledge work” and “process work” is fading as systems handle execution as well as instruction.

Agentic capabilities have also changed management itself. Supervisory tasks, such as allocating work, checking progress and escalating issues, can increasingly be mediated by autonomous systems. In hybrid environments, teams of human and machine agents already collaborate across shared dashboards, with AI handling workflow triage and exception management.

The compounding effect of all three new capabilities

Each of these capabilities is powerful in and of itself, but taken together, their power is compounded. Multimodality provides richer feedback, reasoning improves an agent’s decision quality and agency gives it control over the environment. The result is a self-reinforcing system that continuously improves through interaction. 

This is why job exposure may depend less on the strength of any single capability and more on what happens when they combine. A system that can see, think and act supports far richer work than one that only generates content. That points to AI creeping into practical, everyday tasks, the kind that involve planning, sequencing or inspection rather than pure knowledge.

For example, in 2023, plumbing was a job that few thought would see even a minimal amount of AI automation. However, a multimodal reasoning agent today could notice a damp patch on a wall, infer a leaking joint, draft a repair plan and even generate an invoice or parts list. The plumber still fixes the pipe, but the inspection, diagnosis and supportive actions that lead up to or follow it can increasingly be assisted by AI.

Reasoning and perception are beginning to overlap in almost any role that needs situational judgment. Retail store planning, vehicle servicing and energy infrastructure maintenance all rely on visual understanding and procedural thinking. The compounding effect helps the coordination, diagnosis and verification that connects thought with action.

A look at the most and least impacted jobs

Using our recalibrated exposure scores and newly created velocity scores, we can clearly see how these new AI capabilities could impact a wide range of occupation groups, as well as how much more quickly that change could occur than originally anticipated.

We’ve divided the labor market into two categories:

  • Fast-changing, highly exposed jobs. These jobs have exposure scores that are above the average of 39% and velocity scores above the average of 7.

  • Slower-changing, less exposed jobs. These occupation groups have exposure and velocity scores that are below the average of 39% and 7, respectively. However, job groups in both categories have higher exposure than in our original study.
AI impact on 22 job families

By mapping velocity and exposure scores, it’s clear how much—and how quickly—AI could disrupt specific job families.

AI impact on 22 job families

Figure 4 

Source: Cognizant

Bubble size represents the relative number of workers in the job category.

  1. Management
  2. Business and financial operations
  3. Computer and mathematical
  4. Architecture and engineering
  5. Life, physical, and social science 
  6. Community and social service
  7. Legal
  8. Educational instruction and library
  1. Arts, design, entertainment, sports, and media
  2. Healthcare practitioners and technical
  3. Healthcare support
  4. Protective service
  5. Food preparation and serving related
  6. Building and grounds cleaning and maintenance
  1. Personal care and service
  2. Sales and related
  3. Office and administrative support
  4. Farming, fishing, and forestry
  5. Construction and extraction
  6. Installation, maintenance, and repair
  7. Production
  8. Transportation and material moving

The fastest-changing, most exposed occupation groups

The fastest-changing, most exposed occupation groups

Figure 5
Source: Cognizant

One of the most intriguing trends relates to job groups that were already highly exposed but that now are also seeing the fastest change. These include business and financial operations, management and office/administrative support.  All these job groups have seen their average exposure scores leap from a relatively high 14%–21% in 2023 to a stunningly high 60%–68% today. Additionally, all are now experiencing average velocity scores of 11–14, which is between 4 and 7 points above the average.

This heightened velocity is not simply about automating more drudgery. It reflects the evolution of AI from rigid, process-bound automation to flexible, agentic reasoning. While previous technologies like robotic process automation excelled at automating discrete, repetitive tasks, today’s agentic AI orchestrates complex workflows and aligns administrative functions with the strategic goals they are meant to achieve. It thrives on ambiguity and can make decisions outside the narrow confines of a predefined process.

The potential impact of this agentic workflow capability is particularly visible in roles like financial managers and financial and investment analysts. Where an older AI tool might automatically generate a single financial report, a group of modern AI agents could now manage the entire process: identifying the need for the report based on market triggers, pulling and synthesizing data from internal and external sources, performing a preliminary analysis, drafting executive commentary and scheduling a stakeholder review. As a result, financial managers are seeing an exposure score of 84% and a velocity score of 20. (Other jobs within the business and financial operations job group have lower scores, which pull down the average scores .)

Another example can be found in the C-suite. In our original report, CEOs had a projected exposure score of 25% by 2032. The driving force behind this score was AI’s ability to automate individual tasks like preparing financial documents and analyzing data. Today, the CEO exposure score exceeds 60%. C-level tasks like negotiating contracts, recommending policy changes and even implementing those policies are all within scope of AI support.

Whether board members and shareholders would accept these tasks being automated by AI is another question. The issue of accountability is an important one for C-level AI disruption and will likely constrain the degree to which theoretical exposure levels will play out.

Job families
  • Business and financial operations
  • Management
  • Office and administrative support

60%–68%

Average exposure score range

11–14

Average velocity score range

Specialty job sectors—such as healthcare, education, engineering and the social sciences—are seeing a fast uptick in AI-influenced change. Until very recently, AI was primarily targeted at low-level administrative tasks in these sectors. The majority of job tasks were too complex for AI to tackle, and the most important tasks were not appropriate for AI. This kept exposure scores relatively low.

No longer. With AI now able to handle more complex and important tasks—such as condition diagnosis in healthcare and student assessment in education—AI uptake could be swifter and more extensive.

Take the healthcare practitioners job group, which includes a range of occupations, including surgeons, midwives and radiologists. This group’s projected AI exposure has climbed from 10% in our previous research to 39% today, combined with an above-average velocity score of 8.

Family doctors can now use AI to directly enhance diagnosis, explain procedures and discuss test results. There are also options to drastically reduce administrative burdens by automating referrals, recording patient information and coordinating medical staff. As a result, exposure scores in the field have almost doubled from 33% three years ago to 59% today.

The education sector is similarly impacted. Exposure scores have risen from 11% to 49%, with a velocity score of 11. For example, AI can now help teachers prepare course materials, grade students, conduct research and even facilitate classroom discussions.

Legal work posts the highest velocity score of this cohort, at 12. Lawyers, for instance, have seen an extraordinary rise in exposure scores, from 9% in 2023 to 63% today. Interpreting laws and regulations, analyzing probable outcomes of cases and evaluating findings are all now table stakes for AI, which can also act as an assistant for thornier issues like contract negotiations.

Despite their high velocity, these fields also introduce challenges related to regulation, ethics, accountability for outcomes, empathy and the need for human judgment in complex scenarios. This may slow the potential of AI impact in these job sectors.

Job families
  • Healthcare practitioners
  • Educational instruction
  • Legal
  • Engineering and architecture

39%–49%

Average exposure score range

8–11

Average velocity score range

For some occupation groups, high exposure to AI has already been established as the status quo. These are the fields, particularly in the computer and mathematics arenas, that were first recognized as facing significant disruption three years ago.

The change ahead will be slower for these jobs, largely because many core tasks have already been AI-augmented. Consider that in 2023, the computer and mathematical job group had the highest exposure score of any job family, at 32%. Today, with a 67% exposure score, it ranks as second-highest, behind business and financial operations. However, its velocity score of 9, while above the average, is comparatively lower than the 14 seen in business and financial operations. This is largely due to previous versions of AI already automating significant swaths of tasks in the field.

For developers, AI-powered code generation tools like GitHub Copilot are now an expected part of the toolkit. For statistical assistants, platforms that identify trends and generate predictive models are standard. For these roles, the dynamic is shifting from radical disruption to a period in which productivity gains will be reviewed and assessed, with AI embedded as a foundational layer of daily work.

Job families
  • Computer and mathematical

67%

Average exposure score range

9

Average velocity score range

Job families
  • Business and financial operations
  • Management
  • Office and administrative support

60%–68%

Average exposure score range

11–14

Average velocity score range

One of the most intriguing trends relates to job groups that were already highly exposed but that now are also seeing the fastest change. These include business and financial operations, management and office/administrative support.  All these job groups have seen their average exposure scores leap from a relatively high 14%–21% in 2023 to a stunningly high 60%–68% today. Additionally, all are now experiencing average velocity scores of 11–14, which is between 4 and 7 points above the average.

This heightened velocity is not simply about automating more drudgery. It reflects the evolution of AI from rigid, process-bound automation to flexible, agentic reasoning. While previous technologies like robotic process automation excelled at automating discrete, repetitive tasks, today’s agentic AI orchestrates complex workflows and aligns administrative functions with the strategic goals they are meant to achieve. It thrives on ambiguity and can make decisions outside the narrow confines of a predefined process.

The potential impact of this agentic workflow capability is particularly visible in roles like financial managers and financial and investment analysts. Where an older AI tool might automatically generate a single financial report, a group of modern AI agents could now manage the entire process: identifying the need for the report based on market triggers, pulling and synthesizing data from internal and external sources, performing a preliminary analysis, drafting executive commentary and scheduling a stakeholder review. As a result, financial managers are seeing an exposure score of 84% and a velocity score of 20. (Other jobs within the business and financial operations job group have lower scores, which pull down the average scores .)

Another example can be found in the C-suite. In our original report, CEOs had a projected exposure score of 25% by 2032. The driving force behind this score was AI’s ability to automate individual tasks like preparing financial documents and analyzing data. Today, the CEO exposure score exceeds 60%. C-level tasks like negotiating contracts, recommending policy changes and even implementing those policies are all within scope of AI support.

Whether board members and shareholders would accept these tasks being automated by AI is another question. The issue of accountability is an important one for C-level AI disruption and will likely constrain the degree to which theoretical exposure levels will play out.

Job families
  • Healthcare practitioners
  • Educational instruction
  • Legal
  • Engineering and architecture

39%–49%

Average exposure score range

8–11

Average velocity score range

Specialty job sectors—such as healthcare, education, engineering and the social sciences—are seeing a fast uptick in AI-influenced change. Until very recently, AI was primarily targeted at low-level administrative tasks in these sectors. The majority of job tasks were too complex for AI to tackle, and the most important tasks were not appropriate for AI. This kept exposure scores relatively low.

No longer. With AI now able to handle more complex and important tasks—such as condition diagnosis in healthcare and student assessment in education—AI uptake could be swifter and more extensive.

Take the healthcare practitioners job group, which includes a range of occupations, including surgeons, midwives and radiologists. This group’s projected AI exposure has climbed from 10% in our previous research to 39% today, combined with an above-average velocity score of 8.

Family doctors can now use AI to directly enhance diagnosis, explain procedures and discuss test results. There are also options to drastically reduce administrative burdens by automating referrals, recording patient information and coordinating medical staff. As a result, exposure scores in the field have almost doubled from 33% three years ago to 59% today.

The education sector is similarly impacted. Exposure scores have risen from 11% to 49%, with a velocity score of 11. For example, AI can now help teachers prepare course materials, grade students, conduct research and even facilitate classroom discussions.

Legal work posts the highest velocity score of this cohort, at 12. Lawyers, for instance, have seen an extraordinary rise in exposure scores, from 9% in 2023 to 63% today. Interpreting laws and regulations, analyzing probable outcomes of cases and evaluating findings are all now table stakes for AI, which can also act as an assistant for thornier issues like contract negotiations.

Despite their high velocity, these fields also introduce challenges related to regulation, ethics, accountability for outcomes, empathy and the need for human judgment in complex scenarios. This may slow the potential of AI impact in these job sectors.

Job families
  • Computer and mathematical

67%

Average exposure score range

9

Average velocity score range

For some occupation groups, high exposure to AI has already been established as the status quo. These are the fields, particularly in the computer and mathematics arenas, that were first recognized as facing significant disruption three years ago.

The change ahead will be slower for these jobs, largely because many core tasks have already been AI-augmented. Consider that in 2023, the computer and mathematical job group had the highest exposure score of any job family, at 32%. Today, with a 67% exposure score, it ranks as second-highest, behind business and financial operations. However, its velocity score of 9, while above the average, is comparatively lower than the 14 seen in business and financial operations. This is largely due to previous versions of AI already automating significant swaths of tasks in the field.

For developers, AI-powered code generation tools like GitHub Copilot are now an expected part of the toolkit. For statistical assistants, platforms that identify trends and generate predictive models are standard. For these roles, the dynamic is shifting from radical disruption to a period in which productivity gains will be reviewed and assessed, with AI embedded as a foundational layer of daily work.

The slower-changing, less-impacted occupation groups

The slower-changing, less-impacted occupation groups

Figure 6
Source: Cognizant

Our original analysis found that jobs requiring physical effort were largely unaffected by advances in AI. And while this area of the labor market still has some of the lowest relative exposure and velocity scores, there are, nonetheless, significant swings from three years ago.

Construction and extraction, for example, had a rock-bottom exposure score of just 4% in 2023 and was forecast to grow to 7% by 2032; today it’s 12%, with a velocity score of 3. Transportation and material moving exposure has jumped from 6% in 2023 to 25% today (exceeding the 2032 forecast of 15%), with a velocity score of 6. While comparatively lower than highly disrupted industries such as business operations, these swings were considered beyond the realms of possibility three years ago.

Change is most evident where the physical environment is more instrumented, and workflows are easier to codify. For example, transportation inspectors have seen their exposure scores jump from 6% in 2023 to 55% today. AI can now fully automate tasks such as preparing and submitting reports but is also now encroaching into areas like shipment inspection, observing crews for safety and compliance, and recommending remedial procedures.

To see disruption in a more hands-on role, we can look to the role of brickmasons in the construction sector, which has seen a rapid increase in exposure from 3% in 2023 to 20% today. Tasks such as calculating courses, interpreting blueprints and measuring distances can all be readily assisted by AI, particularly due to advances in multimodality and the increased ability to embed AI capabilities into products like glasses and safety visors.

Numerous factors will continue to modify the pace of AI change for physical labor jobs. For example, liability is high, assets are expensive and operations span mixed digital and physical settings. AI will trim friction around planning, routing and inspection, while people will continue to handle setup, oversight and exceptions.

Job families
  • Construction and extraction
  • Transportation and material moving
  • Production

12%–29%

Average exposure score range

3–6

Average velocity score range

Unlike healthcare practitioner roles that involve diagnosis, research and planning, healthcare support roles such as midwives and nursing assistants sit closer to hands-on care, where outcomes hinge on empathy, trust and continuity of care. Exposure scores have seen a notable rise from 5% in 2023 to 29% today, largely driven by AI’s newer abilities to understand and reason about images, but that score is nonetheless below the average and 10 percentage points below colleagues in the healthcare practitioner group.

As with other sectors, there are big differences across roles. Medical assistants see higher than average velocity scores of 13, with exposure scores jumping from 5% in 2023 to 49% today. A driving force for this increase is AI’s ability to connect the dots between different systems. Tasks like authorizing drug refills, logging patients into a clinic, scheduling appointments and contacting other departments and medical facilities are all likely to see considerable encroachment from AI.

For other roles that involve a large amount of physical tasks like cleaning instruments and dressing wounds, exposure scores are comparatively lower, contributing to the job group’s overall lower exposure and velocity scores. For example, nursing assistants and personal care aides will experience slower change. These jobs involve helping patients with their physical needs and performing clinical tasks that demand dexterity and real-time adaptation to changing conditions. These are difficult to codify and harder to automate without compromising safety or dignity.

Job families
  • Healthcare support

29%

Average exposure score range

6

Average velocity score range

Job groups with some of the lowest AI exposure and velocity scores often function in live environments with uncertain triggers, incomplete information and shared accountability. The same job moves from sensing to deciding to executing, often within minutes. Variability is the norm, and success relies on human abilities such as reading subtle cues, exercising calm judgment and building trust. These qualities remain difficult for AI to replicate, keeping exposure lower and velocity modest.

Take occupation groups like installation and repair, whose exposure scores have risen from 4% in 2023 to a comparatively modest 20%, with a velocity score of 5. Protective services and personal care have also seen relatively low gains compared with the labor market overall, with velocity scores of 6 and 5, respectively.

Automotive mechanics saw their exposure scores spike from just 2% in 2023 to 17% today. AI can help mechanics run through checklists and diagnostics, plan work, review work orders and even support visual inspections. But AI has a much smaller role to play in conducting repairs and installing new parts.

AI is making the most headway behind the scenes. Sensors, cameras, meters, wearables and telematics elevate situational awareness. Coordinators can see location, status, risk and policy in one view, and can draw on playbooks and past fixes in context. Rostering and routing place the right skills in the right location, and reports assemble evidence captured on scene. But the decisive moment still belongs to the technician, officer or care provider, who must weigh context, consent and safety.

Job families
  • Installation, maintenance and repair
  • Protective services
  • Personal care

20%–29%

Average exposure score range

5–6

Average velocity score range

Job families
  • Construction and extraction
  • Transportation and material moving
  • Production

12%–29%

Average exposure score range

3–6

Average velocity score range

Our original analysis found that jobs requiring physical effort were largely unaffected by advances in AI. And while this area of the labor market still has some of the lowest relative exposure and velocity scores, there are, nonetheless, significant swings from three years ago.

Construction and extraction, for example, had a rock-bottom exposure score of just 4% in 2023 and was forecast to grow to 7% by 2032; today it’s 12%, with a velocity score of 3. Transportation and material moving exposure has jumped from 6% in 2023 to 25% today (exceeding the 2032 forecast of 15%), with a velocity score of 6. While comparatively lower than highly disrupted industries such as business operations, these swings were considered beyond the realms of possibility three years ago.

Change is most evident where the physical environment is more instrumented, and workflows are easier to codify. For example, transportation inspectors have seen their exposure scores jump from 6% in 2023 to 55% today. AI can now fully automate tasks such as preparing and submitting reports but is also now encroaching into areas like shipment inspection, observing crews for safety and compliance, and recommending remedial procedures.

To see disruption in a more hands-on role, we can look to the role of brickmasons in the construction sector, which has seen a rapid increase in exposure from 3% in 2023 to 20% today. Tasks such as calculating courses, interpreting blueprints and measuring distances can all be readily assisted by AI, particularly due to advances in multimodality and the increased ability to embed AI capabilities into products like glasses and safety visors.

Numerous factors will continue to modify the pace of AI change for physical labor jobs. For example, liability is high, assets are expensive and operations span mixed digital and physical settings. AI will trim friction around planning, routing and inspection, while people will continue to handle setup, oversight and exceptions.

Job families
  • Healthcare support

29%

Average exposure score range

6

Average velocity score range

Unlike healthcare practitioner roles that involve diagnosis, research and planning, healthcare support roles such as midwives and nursing assistants sit closer to hands-on care, where outcomes hinge on empathy, trust and continuity of care. Exposure scores have seen a notable rise from 5% in 2023 to 29% today, largely driven by AI’s newer abilities to understand and reason about images, but that score is nonetheless below the average and 10 percentage points below colleagues in the healthcare practitioner group.

As with other sectors, there are big differences across roles. Medical assistants see higher than average velocity scores of 13, with exposure scores jumping from 5% in 2023 to 49% today. A driving force for this increase is AI’s ability to connect the dots between different systems. Tasks like authorizing drug refills, logging patients into a clinic, scheduling appointments and contacting other departments and medical facilities are all likely to see considerable encroachment from AI.

For other roles that involve a large amount of physical tasks like cleaning instruments and dressing wounds, exposure scores are comparatively lower, contributing to the job group’s overall lower exposure and velocity scores. For example, nursing assistants and personal care aides will experience slower change. These jobs involve helping patients with their physical needs and performing clinical tasks that demand dexterity and real-time adaptation to changing conditions. These are difficult to codify and harder to automate without compromising safety or dignity.

Job families
  • Installation, maintenance and repair
  • Protective services
  • Personal care

20%–29%

Average exposure score range

5–6

Average velocity score range

Job groups with some of the lowest AI exposure and velocity scores often function in live environments with uncertain triggers, incomplete information and shared accountability. The same job moves from sensing to deciding to executing, often within minutes. Variability is the norm, and success relies on human abilities such as reading subtle cues, exercising calm judgment and building trust. These qualities remain difficult for AI to replicate, keeping exposure lower and velocity modest.

Take occupation groups like installation and repair, whose exposure scores have risen from 4% in 2023 to a comparatively modest 20%, with a velocity score of 5. Protective services and personal care have also seen relatively low gains compared with the labor market overall, with velocity scores of 6 and 5, respectively.

Automotive mechanics saw their exposure scores spike from just 2% in 2023 to 17% today. AI can help mechanics run through checklists and diagnostics, plan work, review work orders and even support visual inspections. But AI has a much smaller role to play in conducting repairs and installing new parts.

AI is making the most headway behind the scenes. Sensors, cameras, meters, wearables and telematics elevate situational awareness. Coordinators can see location, status, risk and policy in one view, and can draw on playbooks and past fixes in context. Rostering and routing place the right skills in the right location, and reports assemble evidence captured on scene. But the decisive moment still belongs to the technician, officer or care provider, who must weigh context, consent and safety.

What this means for the workplace

This velocity of change—and the $4.5 trillion in labor it represents—creates both tremendous opportunity and significant challenges. The question is no longer whether AI will transform work but how quickly organizations can adapt to harness its potential while ensuring their people remain at the center of value creation.

Business leaders and other stakeholders must focus on the following realities:

Our research shows AI’s influence spreading far beyond office environments and knowledge work, reaching into the practical, hands-on parts of the economy. What stands out is how AI is beginning to assist in roles that rely on human perception, context and quick judgment, areas once thought safely beyond automation. 

In these settings, small decisions often determine performance: a technician judging whether a machine is overheating, a driver inspecting a damaged delivery, a nurse assessing a wound. These moments have always relied on experience and the intuition born of that experience rather than formal process. Now, multimodal AI systems capable of interpreting images, sounds and spatial cues can recognize, support and learn from those same judgments. 

This marks a change in how work is understood. Tasks once considered purely manual actually contain embedded cognitive elements that AI can augment. Each small improvement, each instance of better consistency or reduced error, compounds across an organization. When those improvements occur across every shift and every site, the gains become transformative. 

What emerges is a more connected form of work, where digital and physical tasks overlap. The line between knowledge and labor is fading. A warehouse worker using AI to validate product quality, a field engineer guided by an assistive headset, a retail associate capturing store conditions for analysis—all represent a hybrid of physical and digital decision-making. The most human parts of physical work now have the potential for digital enhancement. 

The line between knowledge and labor is fading. The most human parts of physical work now have the potential for digital enhancement. 

Ever since generative AI entered the business scene, important AI developments have seemed to arrive more quickly and more frequently. Organizations need to match their planning and budgeting cycles to this erratic rhythm of change.

Enterprises structured for gradual transformation will, and do, struggle to keep pace. Rigid planning cycles, long budget approvals and fixed technology roadmaps cannot absorb capability shocks of this magnitude. In contrast, organizations with more modular systems, flexible governance and fluid funding flows will show greater resilience. They can test, adopt and redirect resources as the technology moves. 

These organizations are developing what might be called operational elasticity. They expect volatility and have designed around it. A new model release becomes a standard update rather than a strategic crisis. Continuous integration of AI capability becomes the rule, not the exception. 

This also highlights a growing gap between the speed of technological change and the slower tempo of policy and education. Regulatory frameworks, training systems and workforce planning remain tuned to older cycles of industrial adjustment. To remain relevant, institutions will need to build their own adaptive structures capable of learning and reacting almost as quickly as the systems they oversee. 

Organizations with modular systems, flexible governance and fluid funding flows will show greater resilience. They can test, adopt and redirect resources as the technology moves. 

Work and learning are starting to move at the same pace as AI development. As exposure and velocity rise, people must adjust even as the tools they use are still evolving. Adaptability is now an organizational requirement.

The most effective organizations will synchronize the adaptability of their people with the adaptability of their systems. They will create environments where experimentation is part of the job and where feedback between humans and AI tools flows both ways.

Workers are not just using AI but shaping it, testing its limits and redefining their own tasks as they go. Managers need to supervise both people and the agents they use, ensuring that judgment and automation evolve together rather than in conflict.

In sectors such as healthcare, law and education, this interplay is especially visible. AI can now carry out much of the heavy analytical work, yet trust, empathy and ethical discretion remain central. The organizations moving fastest are those that recognize this tension as a source of innovation. They will allow professionals to codesign how AI is applied, preserving the human elements that matter while amplifying what machines do best.

As exposure and velocity rise, people must adjust even as the tools they use are still evolving. Adaptability is now an organizational requirement. 

Skilling must also move at the same velocity as AI itself. In today’s environment, traditional learning cycles operate too slowly. By the time a standard training curriculum is designed and approved, the capability it addresses may have already expanded. Instead, organizations need to treat learning and development as a rapid response mechanism capable of deploying new competencies the moment a technology creates them.

When a new reasoning engine or multimodal agent becomes available, the skilling infrastructure must immediately bridge the gap between the tool’s potential and the employee’s current practice. The focus transfers from broad, role-based certifications to precise, task-based adjustments.

For example, a medical professional does not need to relearn their entire profession when AI diagnostics improve. They need a targeted, immediate adjustment: how to interpret the new agent’s specific output and communicate those findings to a patient. This is a process of constant recalibration. The professional adds a new layer of technical competence while simultaneously doubling down on the human judgment and empathy that the machine cannot replicate.

The organizations that succeed will be those that view skilling as a real-time infrastructure update, ensuring that when the technology jumps forward, the workforce lands right beside it.

Treat learning and development as a rapid response mechanism capable of deploying new competencies the moment a technology creates them.

Getting ready for the $4.5 trillion labor shift

The pace of job change is now deeply interconnected with the rate of AI change. But the two timetables will never completely align. While the shifts our research has revealed reflect the potential of AI as a technology, many other factors will ultimately determine the final outcome.

Regulatory and policy decisions, manager accountability, organizational strategies and workforce adaptability will play a critical role in shaping adoption. Economic conditions, cultural attitudes and ethical considerations may accelerate or slow progress. Finally, breakthroughs or setbacks in AI and related technologies could amplify or diminish the scale and speed of change beyond our current projections.

But given the acceleration we’re seeing, it’s a good chance the next three years will bring even greater change than what we’ve seen in the previous three. Organizations and individuals that invest now in learning, adapting and strategically planning will be positioned to keep pace with AI-driven change and even turn it into a competitive advantage.

Download PDF

For more research, visit us at www.cognizant.com/us/en/insights

Research

Ollie O’Donoghue

Head of Cognizant Research

Duncan Roberts

Associate Director, Cognizant Research

Alexandria Quintana


Senior Manager, Cognizant Research

Ramona Balaratnam

Senior Manager, Cognizant Research

Editorial

Lynne LaCascia
Head of Brand, Thought Leadership
& Research

Marlowe Newman

Director, Thought Leadership & Research

Steven Ulfelder
Senior Manager, Thought Leadership & Research

Mary Brandel

Editor