Psychological Safety in the AI Workplace: A Q&A with the American Psychological Association

As AI reshapes the workplace, one of the most important and often overlooked questions is how workers feel about the change. This Q&A from Partnership on AI, featuring Dennis Stolle, Senior Director of Applied Psychology at the American Psychological Association, explores how psychological safety, defined as a team culture where workers feel comfortable expressing themselves and taking interpersonal risks without fear of retribution, is being directly affected by AI adoption. The APA's Work in America 2024 report found that 41% of workers worry AI may make some or all of their job duties obsolete, and those who hold this worry are significantly more likely to report feeling tense or stressed during a typical workday. For HR professionals, this is not an abstract concern. It is showing up in engagement, trust, and daily performance.

The data on the relationship between psychological safety and AI attitudes is striking. Among workers who reported higher levels of psychological safety, 66% felt confident their employer would retrain them if AI replaced their role. Among workers with lower psychological safety, that figure dropped to 56%. The gap points to something important: whether employees feel capable of having honest conversations about AI at work shapes how they experience the technology. A workplace where concerns cannot be raised safely is one where AI anxiety compounds quietly, rather than being addressed directly.

The generational dimension adds further complexity. Younger workers (aged 26 to 43) are more than four times as likely to use AI at least monthly compared to workers aged 65 and older (44% vs 11%). Yet those same younger workers are more worried about job obsolescence: 48% of workers aged 17 to 25 expressed concern about AI making their role redundant, compared to just 30% of workers 65 and older. Stolle's point is clear: employers need to be intentional about addressing differing AI work styles and the fears that come with them, particularly among the youngest employees who are both the most engaged with AI and the most anxious about it.

Stolle's practical recommendations are directly applicable for HR professionals and people leaders designing AI adoption strategies. Organisations should involve employees in the creation of AI use policies, not just communicate them downward. Transparent communication about how and why AI is used builds trust. AI monitoring tools, in particular, risk eroding psychological safety if employees feel their privacy is compromised without good reason. And well-designed AI tools, by reducing repetitive tasks and freeing up time, can actually create more space for the collaborative, human-to-human interaction where psychological safety is built. The technology is not the problem. How it is introduced, governed, and talked about is what determines whether people feel safe or threatened.