American workers who avoid artificial intelligence (AI) may find themselves on thinner ice than their tech-forward peers. According to recent data from Gallup, professionals who rarely or never utilize AI tools are significantly more likely to face layoffs than those who integrate them into their regular workflows.
The research highlights a growing divide: 62% of laid-off workers identified as AI non-users (using the technology once a year or less). In contrast, only 50% of currently employed workers fall into the non-user category.
The AI Usage Divide
The frequency of AI adoption appears to correlate directly with job security. Gallup’s data reveals a distinct pattern when comparing currently employed workers to those who have been recently laid off:
Frequent Users: 28% of currently employed workers use AI daily or weekly, compared to just 22% of laid-off workers in their previous roles.
Infrequent Users: 22% of employed workers use AI a few times a month or year, while only 16% of laid-off workers fit this description.
The Vulnerability Factor: "This pattern holds even after accounting for age, education, industry type, and the time elapsed since the layoff," Gallup noted, suggesting that a lack of AI adoption genuinely increases market vulnerability.
Tech Sector Workers Hit Hardest
While the trend spans multiple industries, the impact is most severe in the technology sector. Tech workers who used AI monthly or less were three times more likely to be laid off (18%) than their peers who leveraged AI at least monthly (6%).
Because the tech sector has faced an elevated baseline of workforce reductions over the last few years, the gap between AI adopters and non-users in this field is particularly stark.
Is AI Replacing Jobs, or Just Restructuring Them?
Despite the correlation, workers aren't necessarily pointing the finger at algorithms for losing their jobs. When Gallup asked laid-off respondents to pinpoint the primary reason for their termination, only 1% blamed AI or automation directly.
Instead, respondents cited more traditional corporate adjustments:
Organizational restructuring and downsizing: 15%
Role elimination: 3%
The Takeaway: While employers aren't explicitly telling workers they are being replaced by AI, the technology is almost certainly driving behind-the-scenes restructuring. Business leaders are actively re-evaluating workforce structures, prioritizing employees who can leverage AI to boost productivity, and downsizing roles that remain tethered to legacy processes.
As corporate layoffs hover around 21% as of early 2026, the data sends a clear message to the workforce: staying competitive in modern industry is less about being replaced by AI, and more about being replaced by someone who knows how to use it.
How AI is Powering Autonomous Robot Workers—for Factories, Warehouses, and Eventually Homes
Top robotics researchers and founders explain the rapid evolution of robot autonomy.
Agility Robotics' Digit humanoid robots are already handling tasks in warehouses and on factory floors. Credit: Agility Robotics
Self-driving robotaxis navigating city streets and delivery drones zipping through the skies have made one thing clear: autonomy is transforming how machines interact with the physical world. The next frontier? General-purpose robots that can assist humans across workplaces and, one day, homes—powered by advanced AI.
This vision has drawn top researchers into entrepreneurship and attracted billions in investment. Yet it builds on decades of incremental progress. “When I started about 15 years ago, autonomy meant getting a robot from point A to point B,” said Matt Malchano, vice president of software at Boston Dynamics. “Now, we think in terms of a huge space of tasks that robots can perform on their own.”
Early robotics struggled with basics like navigation and balance. The 1979 Stanford Cart took five hours to travel 20 meters. The first self-balancing bipedal robot appeared in 1996. Today, the International Organization for Standardization defines robotic autonomy as the ability to perform intended tasks based on current state and sensing, without human intervention.
Modern AI—particularly reinforcement learning from the 2010s and large foundation models from the 2020s—has dramatically accelerated progress. These tools enable robots to handle sequences of activities, understand tasks, and adapt to unpredictable environments.
The AI Revolution in Robotics
Sergey Levine, UC Berkeley computer scientist and cofounder of Physical Intelligence, emphasizes that the future lies in versatile AI models rather than a single “super humanoid.” “I don’t think it will be the one ultimate robot,” he said. “It will be a general AI model that can power lots of different robots well-suited for their jobs”—from compact arms in small apartments to rugged machines on farms.
Achieving true open-world autonomy requires major leaps in perception, motor skills, error recovery, instruction following, and generalization. Researchers combine reinforcement learning (trial-and-error training, often in simulation or the real world) with large pre-trained models that provide common-sense priors from vast datasets of images, text, and video.
“Reinforcement learning is like practicing your tennis swing many times,” Levine explained. “But you first need basic common sense to get started.” Teleoperation—humans remotely guiding robots—supplies valuable training data, while “world models” help robots predict action outcomes. Still, a significant data gap remains. Simulations are cheap but miss real-world messiness; real-world trials are expensive and slow.
Current systems excel at specific tasks in controlled conditions or show broader but less reliable capabilities. “We really want something extremely good at all things,” Levine noted, “and that’s still at the frontier of research.”
From Labs to Real Work
Specialized robots already deliver value today. Boston Dynamics’ Spot quadruped conducts autonomous inspections in hazardous environments, such as converter stations and highway culverts. Its wheeled Stretch robots handle boxes in warehouses for companies like DHL, adapting to varied loads and truck configurations through real-world experience.
The company’s electric Atlas humanoid is now scaling up, with plans for deployment at Hyundai’s Georgia EV plant by 2028. Leveraging Hyundai’s manufacturing expertise aims for 30,000 units annually. Early labor concerns have emerged, but the focus remains on proving reliability. “We’ve come to expect that if you ask a person to do a task, they’ll do it right... almost all the time,” Malchano said. “We’re still understanding what it takes to achieve that level for general-purpose, AI-driven robots.”
Agility Robotics has taken a leading commercial role, deploying its Digit humanoids in warehouses (GXO, Amazon pilots), automotive lines (Toyota, Schaeffler), and e-commerce facilities. The robots have logged over 65,000 operational hours, primarily moving totes. Next steps include handling diverse items, boxes, and less structured retail backrooms. Co-founder Jonathan Hurst sees a gradual path: warehouses → retail → delivery → homes, potentially spanning decades.
Hurst cautions against overhyping near-term home robots. “Companies that claim their humanoid robots will be safely working inside homes... are either lying or wrong.” True home autonomy must handle unpredictable scenarios, like receiving a baby from a person.
Safety First
Safety remains the paramount hurdle. The 1979 death of a Ford worker crushed by a robotic arm underscored the risks. Agility deploys Digit in isolated “work cells” for now. Its upcoming Digit v5 aims to be the first “AI-enabled, cooperatively safe humanoid,” capable of detecting people, stopping, and safely crouching.
Both Agility and Boston Dynamics contribute to ISO safety standards for industrial mobile robots. In surgery, where the stakes are highest, autonomy is tightly constrained. Intuitive’s systems remain primarily human-driven, with levels of assistance. Higher autonomy levels (3–5) promise precision and consistency but require rigorous validation for complications like bleeding or anatomical variations.
“The question is how much decision-making and action execution we are delegating to machine versus human,” said Bhushan Patel of Intuitive.
A Self-Fulfilling Prophecy
The AI boom has supercharged interest in robotics, drawing thousands of talented engineers. Dipam Patel, a Purdue PhD student and Army Research Lab researcher, works on robots for search-and-rescue—traversing rubble, manipulating obstacles, and achieving whole-body control. Challenges include catastrophic forgetting in training, onboard compute limits, and multi-step planning.
Patel shares Levine’s pragmatism: “People are like, ‘we need a human-like robot,’ but we don’t really need that. We just need a robot that can do stuff.”
Hurst believes this influx of talent creates momentum. With motivated engineers dedicating careers to the field, the gradual, hard-won advances in embodied AI will compound—bringing autonomous helpers from factories to homes closer to reality, one reliable task at a time.