A comprehensive research study conducted by scientists from institutions across the United States and United Kingdom has uncovered troubling cognitive effects associated with regular artificial intelligence use. The findings, detailed in a paper titled "AI assistance reduces persistence and hurts independent performance," suggest that while AI tools enhance immediate task completion, they exact a significant neurological cost on users.
The research team discovered that just ten minutes of AI assistance can create technology dependency, subsequently leading to diminished performance and elevated burnout once the tools become unavailable. The study focused specifically on individuals relying on AI for reasoning-intensive cognitive tasks—including writing, software development, and creative brainstorming—among the most prevalent real-world applications.
In the initial experiment, 350 participants were assigned fraction-based mathematical problems. Half received access to a specialized chatbot powered by GPT-5 technology, while the control group worked independently. When AI access was abruptly removed midway through the assessment, the previously AI-assisted group experienced sharp performance drops accompanied by widespread task abandonment. A larger replication involving 670 participants yielded identical patterns. A subsequent trial using reading comprehension questions produced comparable results across all metrics.
The data revealed a particularly concerning behavioral shift: participants didn't simply produce incorrect answers after losing AI access—they fundamentally lost willingness to attempt problems without technological assistance. According to Rachit Dubey, an assistant professor at the University of California and study co-author, "People's persistence drops" significantly once AI tools vanish from their workflow.
The implications extend beyond individual productivity concerns. Dubey cautioned that widespread AI integration in educational settings risks creating "a generation of learners and people who will not know what they're capable of," potentially undermining foundational skill development and self-reliance among students. These findings arrive as educational institutions worldwide grapple with AI integration policies and debate appropriate deployment strategies that balance technological advantages against cognitive development priorities.