New TD Economics research suggests AI is more likely to help people work more efficiently than replace them, while human expertise, judgment and creativity remain essential.
Key Points:
- AI is more likely to change how people do their jobs than replace large numbers of workers.
- Human judgment, experience, creativity and trust remain critical in many jobs, even as AI takes on more routine tasks.
- For most businesses, AI works best alongside employees. The costs and complexities of automation mean companies often see greater value in using AI to support workers rather than replace them.
From drafting emails and summarizing documents to helping analyze information, AI is already changing how many Americans work. But according to new TD Economics research, the technology is more likely to become a tool that helps people do their jobs more efficiently than a force that replaces large numbers of workers.
Even when AI can automate a task, that doesn't mean the technology is reliable or cost-effective enough for an organization to use it to replace large numbers of workers, according to the report authors Rannella Billy-Ochieng’, Senior Economist, and Thomas Feltmate, Director and Senior Economist at TD Economics.
Instead, AI's effects on the job market depend on three main factors: what the technology can reliably do, whether automation makes financial sense and how quickly U.S. companies adopt it.
Why AI adoption doesn't equal a looming "job apocalypse"
A 2026 TD AI Insights survey found that 83% of the 2,500 Americans surveyed use AI-powered tools and applications to support their work. Despite more comfort with using these tools, survey respondents said human involvement and judgment are still critical in their workplaces. AI can make recommendations, but human oversight is essential.
Sixty-eight percent of the respondents were also less worried about AI fully taking over their job than they were one year ago, and 91% believed that AI should be used in the workplace.
"The future of work isn't about humans versus AI," said Ted Paris, Head of Analytics, Intelligence & AI, TD Bank U.S. "It's about how people and technology work together to unlock productivity, innovation and better outcomes."
Rannella and Thomas said that a large-scale, AI-driven “job apocalypse” would require several conditions to happen at once — an unlikely event in the near term. These include the ability of AI tools to accurately and completely automate decisions, firms' capacity to fund these automated processes and overcome operational hurdles, and how quickly AI is fully adopted.
Why humans still play a critical role in the workplace
For many workers, AI increasingly functions as a digital assistant, helping summarize meetings, organize information, draft communications and automate repetitive administrative tasks. While the technology can save time, people still provide the judgment, creativity and relationship-building skills that many jobs require.
Larger job losses could happen if and when AI can consistently and accurately perform tasks with little human oversight. While today's leading AI models are making impressive advances, particularly in areas like software coding and development, many jobs still rely on human judgment, experience and interpersonal skills.
Research suggests that fully autonomous AI remains rare — and that AI is more likely to change how we work than fully replace people. That distinction is important because most jobs are made up of many different tasks. Automating one of those tasks doesn’t necessarily mean AI can replace the person doing the job. In fact, one study found that just 1.8% of jobs today could have more than half of their regular tasks fully automated with current AI and software tools.
AI must become more cost-effective
Just because AI can perform a task doesn't mean it makes financial sense for a business to automate it. Beyond the technology itself, companies must invest in implementation, employee training, compliance, cybersecurity and ongoing oversight. Those costs can add up quickly. In fact, one MIT study that investigated the relative cost of deploying AI on tasks using computer vision systems found that while 36% of occupations included at least one task that could technically be automated, only 8% included a task that was economical to automate at scale. In other words, being able to automate a task doesn’t necessarily mean it makes financial sense to do so.
AI can deliver savings on routine work such as summarizing documents, drafting emails, and extracting information. But costs rise significantly when tasks require complex reasoning, planning or decision-making. At the same time, businesses face uncertainty around the future cost of AI infrastructure, computing power and specialized hardware.
The result: while AI adoption continues to grow, the current business case for replacing large numbers of workers is much weaker than the case for using AI to help humans work more efficiently.
AI adoption would have to greatly accelerate
While consumers have embraced AI quickly, regular adoption across businesses has been much slower. Many companies are still experimenting with the technology, and day-to-day use largely depends on the industry.
That slower pace matters. AI is increasingly being used to support existing workflows rather than replace them, giving workers and businesses more time to adapt. Economic constraints, implementation challenges and industry-specific needs are all likely to slow widespread adoption.
Labor markets are constantly evolving, with jobs created and eliminated every year as businesses grow, shrink and adopt new technologies. Some job impacts have occurred, especially in positions that are easier to automate.
Currently, many employers are adjusting by scaling down hiring instead of implementing widespread layoffs. According to the report, this adjustment is primarily affecting younger workers entering the job market, who are finding fewer entry-level opportunities.
For AI to drive large-scale job losses, adoption would need to happen much faster than these normal adjustment processes. So far, employment trends suggest that transition remains gradual rather than disruptive.
At the same time, AI is creating demand for new skills and careers. Growth in AI infrastructure, data centers and related industries is already generating jobs, showing how technological change can create new work even as it transforms existing ones.
"The ultimate employment impact depends not just on the jobs displaced, but also on the economy’s ability to create and absorb workers into new ones," noted Rannella and Thomas in their report.
What this means for the American workforce
AI's capabilities and impacts change nearly every day, so concerns about how it will affect the job market aren't likely to end until more is known about i's true feasibility as a work tool.
For now, the evidence suggests AI is more likely to reshape how people work than eliminate large numbers of jobs. As businesses continue experimenting with the technology, its effect on the workforce will depend in part on how people learn to use it and how effectively the economy creates new opportunities.
The bottom line for future of work: "There will be a gradual change, rather than a rapid and significant disruption," said Rannella and Thomas.
