Atlassian Teamwork Lab Study Reveals Psychological and Operational Impacts of AI Dependency Through "Play/Pause" Experiment

For a significant portion of the modern knowledge-based workforce, the integration of artificial intelligence into daily workflows has transitioned from a novel experiment to an automated reflex. As these tools become increasingly ubiquitous, the Teamwork Lab at Atlassian sought to quantify the qualitative shifts in how work is experienced and executed when AI is present versus when it is removed. To investigate the depth of this integration, the research group conducted a controlled experiment titled "AI Play/Pause Days," designed to isolate the impact of AI on productivity, stress, and professional discernment. The central question of the study was simple yet provocative: what happens to a professional’s performance and psychological state when AI is intentionally removed for a full workday?

The experiment, conducted on June 9 and 10, 2026, involved approximately 123 talent acquisition professionals at Atlassian. The methodology was structured as a two-part comparative study. On the first day, designated as "Play Day," participants were instructed to maximize their use of AI for every possible task within defined safety and operational guardrails. These tools included Atlassian’s proprietary AI-powered teammate experience, Rovo, and the research assistant, Dia, alongside external tools such as Google Gemini and various built-in AI features within their existing software stack. On the second day, designated as "Pause Day," the same participants were required to refrain entirely from using AI tools, returning to traditional manual processes for all work-related tasks. Throughout the experiment, participants recorded their observations via Loom video reflections and completed surveys to capture both quantitative data and qualitative sentiment.

The results of the study indicate that AI has become so deeply embedded in the modern workday that many professionals no longer notice its contributions until the technology is withdrawn. The "Pause Day" revealed a significant psychological dependency on these tools. According to the data, 82% of participants reported a recurring urge to "just ask AI" when faced with a problem, highlighting the reflexive nature of current AI usage. Furthermore, the absence of AI created a palpable sense of friction; 47% of participants characterized their daily tasks as "uncomfortably hard" without technological assistance. Perhaps most notably, 46% of the talent acquisition professionals reported that their non-AI-assisted skills felt "rusty," suggesting that the long-term use of automation may lead to a perceived or actual atrophy of traditional professional competencies. Nearly one in four participants admitted to reaching for AI tools out of sheer habit rather than a specific functional need.

In contrast, the "Play Day" data provided a clear picture of the benefits associated with high AI integration. Participants reported that their priority work felt 44% easier to complete when AI was fully utilized. The psychological benefits were also measurable, with reported stress levels dropping by 22%. These metrics suggest that AI effectively lowers the "activation energy" required for complex tasks, allowing workers to navigate their responsibilities with greater perceived ease and less emotional strain. However, the use of AI was not without its own set of challenges. In the Loom video reflections recorded during the "Play Day," 44% of participants flagged the "verification cost" associated with AI. This refers to the time and cognitive effort required to check, validate, and correct the outputs generated by AI tools. This finding underscores the fact that while AI reduces the effort of creation, it introduces a new layer of responsibility in the form of oversight and quality control.

The experiment also tracked the emotional and professional sentiment of the participants using specific descriptors. The "Play Day" and "Pause Day" yielded different sets of top adjectives, though they shared one commonality. For the "Play Day," the top three adjectives reported by participants were "Engaging" (64%), "Satisfying" (51%), and "Purposeful" (47%). On the "Pause Day," the sentiment shifted toward "Slow" (50%), "Challenging" (39%), and "Purposeful" (34%). The fact that both days were rated as "Purposeful" is a point of interest for the researchers. On the "Play Day," the sense of purpose came from the opportunity to master new tools and explore the frontiers of AI-assisted productivity. On the "Pause Day," the sense of purpose was derived from the struggle itself; the challenge of manual work helped participants identify exactly where AI was providing genuine value and where it was merely serving as a habit-based substitute for effort.

Allison White, Lead Behavioral Scientist at Teamwork Lab, noted that the "Pause Day" served as a vital diagnostic tool for the organization. White stated that turning the technology off helped people notice the specific areas where AI provided significant help versus the areas where its value was overestimated. The contrast between the two days highlighted the difference between productive AI use and indiscriminate AI use. The study suggests that without such pauses, workers may lose the ability to discern when the "verification cost" of AI outweighs the benefits of its initial output.

One of the most significant findings of the experiment involved the social and collaborative dynamics of the team. While the intensity of AI use significantly altered how individuals felt about their personal workloads, it did not result in any reported change in how much they collaborated with their colleagues. AI, in this context, appeared to be an individual productivity multiplier rather than a replacement for human-to-human interaction. However, the most profound impact on the team occurred after the two-day experiment concluded. When the talent acquisition teams met to share their reflections and compare notes on their experiences, 89% of participants reported that their team had developed a clearer, shared understanding of when to use AI.

This shift from individual insight to "shared judgment" is a key takeaway for organizational leaders. The Teamwork Lab report suggests that when a single team member learns to use an AI tool effectively, it creates an isolated pocket of efficiency. However, when a team collectively reflects on the strengths and weaknesses of the technology—as they did following the Play/Pause experiment—that individual knowledge is transformed into an updated set of work practices for the entire group. This collective discernment allows teams to avoid the trap of becoming overly dependent on AI for tasks that might be better handled through human expertise.

For managers and organizational leaders, the study offers a strategic framework for refining AI adoption. The data suggests that while AI makes work feel easier and less stressful, the risk of "indiscriminate use" can lead to overall efficiency losses due to the aforementioned verification costs. The experiment concluded that taking a periodic break from AI is a simple and effective method for assessing a team’s level of dependency. Rather than fearing that AI will permanently undermine human capabilities, the Teamwork Lab suggests that the real risk lies in a lack of discernment. By participating in a "Pause" day, teams can regain confidence in their own expertise and recalibrate their relationship with their digital tools.

Following the experiment, the participants overwhelmingly reported a higher degree of confidence in knowing when not to use AI. They expressed a renewed trust in their own professional expertise, having successfully navigated a workday without technological assistance. The Teamwork Lab has since integrated this methodology into the "Atlassian Team Playbook" as the "AI Play/Pause Day Play," offering it as a tool for other organizations to assess their own "AI discernment."

Ultimately, the study by Atlassian’s Teamwork Lab reinforces the idea that the value of AI is best realized when its use is intentional rather than habitual. While the technology offers undeniable gains in ease and stress reduction, the "Pause" day experiment demonstrated that a temporary return to manual processes can sharpen a team’s focus, clarify the true utility of their tools, and ensure that human judgment remains the primary driver of professional excellence. The findings suggest that to truly refine the use of AI at work, teams must occasionally be willing to turn it off.