The Great AI Attrition: Why your best AI talent is eyeing the exit

The Great AI Attrition: Why your best AI talent is eyeing the exit

Top AI performers are twice as likely to quit, according to new research from the Upwork Research Institute. As leaders focus on using AI to maximize worker productivity, they're losing sight of the flight risk created across the very workers using AI most effectively. So what are the implications of an en masse AI talent drain?

We need only look as far back as the 1990s for clues.

Today’s AI efficiency experts are like the early adopters of the burgeoning internet. In the ‘90s, self-taught employees started building websites, automating workflows, and reshaping communication at work. Instead of being seen as providing a core business advantage, these innovators were often viewed through a narrow lens, tasked with supporting systems rather than reinventing them. Because they weren’t seen as valuable, early adopters left—flocking to start-ups and forward-thinking companies that were building a new way of working.

Those finding the most productivity with AI today are following the same path. They’ve just begun to unlock the strategic advantage of AI in their day-to-day work and are primed to spend reclaimed time on deep work, problem solving, and creativity. But these gains, while tangible, are misaligned with the static direction of their organizations.

Just as the internet vanguard in the 1990s found their contributions undervalued, today’s AI pioneers are seeing their newfound ability to scale forced to be spent on outdated, linear work models. Instead of being given the trust to solve old problems in new ways, they’re met with an expectation to meet more of the same old metrics.

The data highlights this disconnect between AI workers and leadership, as these workers are 32% more likely to say they have no idea how their company will achieve its AI goals. So it’s no surprise that this lack of visibility and transparency is causing frustration and leading them to leave.

These workers have reached a breaking point and are voting with their feet. They’re looking for organizations that are actively redesigning the work experience to integrate AI, rather than viewing the culture shift as an afterthought.

The Upwork Research Institute calls this the Great AI Attrition. In this article, Upwork, an online marketplace for hiring skilled freelancers, examines what's behind the Great AI Attrition and how to prevent this volatility across the highest performers on your team.

High AI performers face a growing imbalance

In Upwork’s research, high-AI performers—those who report a 40% increase in productivity due to using AI at work—are disproportionately likely to say they’re planning to quit. These are the employees who’ve embraced AI to boost their output and are redefining what it means to get the job done.

To be clear, they’re not just automating tasks. They’re augmenting their work with AI by finding the right mix of tasks to press repeat on, and of tasks that deserve their full attention. For example, graphic designers use AI to speed up the initial concept phase and routine edits, giving them more time to focus on creative strategy and storytelling.

Recent data from Upwork supports this, showing that freelancers are using AI for augmentation 71% of the time and for automation only 29% of the time. AI isn't replacing human thinking; rather, it's enhancing it.

This talent is changing the playbook for what work means and how it’s done. But what they get in return is more work with no meaningful change to how their roles function.

There is a growing imbalance: As innovative AI workers carry the cognitive and creative burden of workplace transformation, they don’t have the support of a matching redesign of their roles, teams, or work cultures.

More output, same operating model

While AI top performers reinvent how they work, they’re operating in the increasingly antiquated systems of yesteryear. Their productivity gains are met with greater work expectations, not more trust. When someone uses AI to complete tasks faster or better, the default response from many organizations is to give them more tasks—not more time to think, create, or lead.

The result? Burnout and disengagement, with 88% of the highest-performing AI users reporting overload and disconnection in a recent study from The Upwork Research Institute.

This begs the question: What do these workers really want? Leadership that understands how AI changes the time, attention, and decision-making equation. They’re looking for workplaces where smart collaboration replaces back-to-back meetings, where creativity is supported by AI, and where human connection remains a priority.

So, is a work redesign on your agenda? If not, here are three ways to start building a human plus AI system that will keep your future-ready talent engaged and thriving.

What leaders can do now: From adoption to integration

Effective organizations—those that Upwork has dubbed Work Innovators—have moved past simply adopting AI tools. They're integrating AI into the entire work experience. They're listening to employee sentiment, redefining roles, and shifting the way we connect, learn, and lead.

To retain top AI talent, start here:

1. Communicate your AI strategy, even if it’s not final

Workers finding the greatest productivity with AI are the most disconnected from the organization. Why? They don’t know where leadership is headed.

The messages they do hear often focus solely on AI efficiency gains—which can result in concern that their jobs are undervalued, or even expendable. Such organizational ambiguity can lead workers to quit before they’re forced out or because they feel that their roles are being changed for them, not with them.

Clear communication around AI strategy is paramount, even if that strategy is evolving. Transparent leadership builds trust and belonging — essential to the very workplace connection required to retain top talent. The bottom line: Prevent this unclear worker swirl with greater visibility into your AI direction.

Here's how: Share a working vision for AI, and be honest about its progress. Point out what's being developed, and invite (and respond to!) feedback. Use recurring forums like all-hands or team-level updates to present direction and company values. Example transparent leadership opener:

AI is changing how we work, and many feel both excited and uncertain about what that means for this organization. While we don’t have all the answers, we’re building AI into our workflows, roles, and how we interact every day.

As the technology is in constant motion, we’ve developed an AI vision that’s a work-in-progress. Know that your input will help shape it. Here’s where it’s headed so far ...

2. Read the room: Listen for early signs of disengagement

AI-burnout is little different from other types of burnout—it quietly shows up through subtle signals. Help managers spot the early warning signs such as fewer meeting contributions, skipped social interactions, and steady output with decreased collaboration.

In other words, tune in to whether this worker sect is pulling away or feeling isolated. And if they are, do something about it. Communication goes a long way in getting a read on and solving most problems.

Here's how: Use pulse surveys, stay interviews, and regular check-ins to track shifts in engagement. Flip the manager script in 1:1s to go beyond "How's work?" to more prescriptive conversation starters like:

  • How's your work experience overall?
  • How's the way we work working for you?

Related to AI:

  • How are you using AI?
  • How do you want to use AI?
  • Are you clear on company expectations of how to use AI?

3. Redesign roles to emphasize innovation

If AI has made someone 30% more efficient, don’t give them 30% more work. Instead, urge workers to use that time to make space for creativity, experimentation, or mentorship. Lean into the innovation that this talent brings.

Acknowledge that most of today’s roles weren’t built for AI augmentation; dust off job descriptions and update them to include the strategic impact current AI-enabled jobs introduce.

For example, a customer support role can evolve from handling routine inquiries to focusing on complex problem-solving and customer retention, as AI triages baseline FAQs.

Here's how: Work with employees who have seen the greatest gains from AI to think about how to update their roles. This shared sense of ownership in establishing role structure will reconnect them to the organization, and allow them to stretch into work no one thought was possible.

4. Help your workforce to learn how to learn

AI integration requires you to onboard a behavioral shift in the way you work, which includes trying and learning new skills. Your top AI talent craves fresh data, methodologies, and room to grow both within AI and their career. Meet them where they are by renewing your focus on coaching—something that will benefit both advanced and beginner AI users.

According to Upwork's 2025 In-Demand Skills Report, career coaching, training, and development are among the fastest-growing skills. This is a powerful signal: The more we automate with AI, the more workers look for human-centered guidance.

Here's how: Retrain your workforce to "learn how to learn" by fostering a culture of human coaching at your organization. Encourage experimentation and allow for slack in the system. This test-and-iterate approach is where progress happens.

Turn this retention problem into a reinvention moment

Just like the early internet era, we’re seeing the emergence of a new class of worker—the AI-native professional who expects work to function differently, and systems to evolve. Leaders who recognize and act on this shift will not only get ahead of The Great AI Attrition, but also usher in a new level of human-AI performance and creativity.

This story was produced by Upwork and reviewed and distributed by Stacker.