For many knowledge workers, reaching for AI tools is now automatic. But the Teamwork Lab wanted to know if using AI actually changes how work feels and how much gets done.
So Teamwork Lab tested: What happens when you take away AI for a day?
research at a glance
What we did:
- In June, Teamwork Lab invited around 250 Talent Acquisition team members to experience both an AI Play day (use AI for everything you can) and an AI Pause day (regular work with no AI tools).
- Each participant experienced both conditions and served as their own comparison. The lab analyzed daily survey data from 123 people; 96 completed the post-experiment survey and recorded a brief Loom video, using Atlassian’s async video tool.
What we learned:
- AI use is automatic. 82% felt the urge to “just ask AI” on the Pause Day.
- Shared reflection helps teams refine judgment. People didn’t report working more or less with each other depending on whether they were using AI. The difference is that after comparing notes, 89% reported a clearer shared understanding of when to use AI.
- The real opportunity is developing discernment. After the experiment, participants were overwhelmingly more confident about when not to use AI.
When AI is always on, you stop noticing when it helps
Because the survey participants are regular users of Rovo and Dia – Atlassian’s AI-powered teammate experience and research assistant – Teamwork Lab was able to isolate what actually changes when you take away the use of those tools.
The lab discovered that AI use is so deeply embedded in the workday that participants reach for it almost without a second thought. When asked to pause, 82% felt the urge to “just ask AI.” 47% said tasks felt uncomfortably hard. 46% said their non-AI skills felt rusty, and nearly one in four reached for it out of habit.
On the Play day, AI use had real value. Participants reported that priority work felt 44% easier, and stress was 22% lower. However, 44% of the Loom reflections did flag the verification cost of checking and fixing AI output.
| Top 3 adjectives for Play day | Top 3 adjectives for Pause day |
|---|---|
| Engaging (64%) Satisfying (51%) Purposeful (47%) | Slow (50%) Challenging (39%) Purposeful (34%) |
Interestingly, participants found both days to be “Purposeful”. On the Play day, people set aside time to get familiar with new AI tools. While participants found the Pause day challenging, it also helped them figure out where AI really was helping, and where it wasn’t.
Pause day provided a stark contrast to both Play day and the typical way we work right now. Turning it off helped people notice where AI really helped, and where it wasn’t providing as much value as they thought.—Allison White, Lead Behavioral Scientist, Teamwork Lab
Shared reflection on AI use helps teams improve judgment
While the intensity of AI use changed how people felt about their own work, people did not report any change in how much they worked with their colleagues.
The most significant difference came after the Play/Pause days had concluded. When teams came together to compare notes on what each person learned, 89% said their team now had a clearer shared understanding of when to use AI.
→ Use the AI Play/Pause Day Play to assess your own team’s AI discernment.
When one team member decides to learn how to use an AI tool, that creates individual insight. Sharing that insight with their team members turns that knowledge into shared judgment and updated work practices.
To refine AI use at work, take a break
For managers, knowing when to use AI and when not to is a valuable skill. AI makes work feel easier, less stressful, and more productive, but it has a verification cost if your team is using it out of habit.
After the Play/Pause experiment, participants overwhelmingly said they felt more confident about when not to use AI and to trust their expertise more.
- 92% of participants left more confident about when not to use AI.
- 83% planned to change how they use AI day to day.
- 86% would recommend the exercise to other teams.
If you’re worried about whether your team is becoming dependent on AI, taking a pause is a simple and effective way to decide. The real risk isn’t that AI will undermine your team’s capabilities, but it is possible that indiscriminate use can affect their overall efficiency. With a Play/Pause break, teams can learn where AI really helps, and where it doesn’t.
Methodology
On June 9 and 10, 2026, Teamwork Lab conducted a two-day experiment with approximately 123 Atlassian talent acquisition professionals. On the first day, participants were instructed to use AI for everything they can in their work, within defined guardrails (Rovo, Dia, Gemini, built-in AI features). On the second day, participants were instructed to refrain from using AI in their work. At the end of each day, participants reflected on their experience by recording a brief Loom video and completing a brief survey. Participants also completed a post-experiment team reflection and survey.
