Supporting safe and open AI use (activity)

Ongoing, with regular check-ins | Leader-led, with team participation | Moderate effort

Use this activity to:

• create clear boundaries for safe AI experimentation
• encourage open discussion
• turn shared learning into better ways of working.

This activity supports our guidance on how to navigate change.

Start with a team session, then revisit it through planning, reflections and regular check-ins.

Safe and open AI use helps staff learn through practice rather than guesswork. It reduces hidden risks and builds shared confidence as AI use expands over time.

To support your team:

• make experimentation safe, not hidden, so they can learn by trying ideas openly
• set clear boundaries so they understand what safe use looks like
• encourage questions and discussion without judgement or penalty
• support learning over time through check‑ins and shared reflection.

A safe environment helps staff move from early experimentation to practical and creative AI use that delivers real value.

Bring your team together to discuss if they already use AI, where it could help, and where uncertainty remains.

Start by agreeing on how staff can experiment safely and openly. Keep the discussion practical – focus on real tasks, real tools and real work.

Work through the following stages to help your team build safe and open AI use over time. 

At this stage, focus on making it safe to talk about low-risk AI use.

Start by creating a shared understanding that learning, curiosity, and small-scale AI use are encouraged within clear and simple boundaries.

Do one practical thing to show that safe exploration is supported, such as:

• tell your team that low-risk AI use is acceptable when it only involves public or non-confidential information
• share a few safe everyday examples, including non-work uses, to normalise learning and curiosity
• make space in a team meeting for staff to share what they have tried
• remind your team that early use is about learning what works, not performance or delivery.

The goal at this stage is simple: make it feel safe for staff to ask questions, test ideas and talk openly about what they are trying.

This stage focuses on helping your team understand what safe and effective AI use looks like over time.

Build a shared understanding of the behaviours and supports that help staff use AI safely and confidently at work.

Focus the discussion on:

• what makes AI use feel safe rather than risky
• what helps staff share learning without fear of getting it wrong
• how clear boundaries can increase confidence to try new things.

Ask your team for practical examples outside of work, such as planning trips, researching purchases or learning new skills.

Talk about what good habits look like in these examples, then explore how the same habits could support work and team processes. This helps the team move to shared confidence and better day-to-day habits.

Next, put simple structures in place so staff can reflect on, practice and share what they learn from using AI.

It helps to:

• complete the AI screening questions
• run an activity for checking team readiness
• create a simple way for your team to share useful lessons in meetings or team channels.

Work to turn safe AI use into strong, long-term habits.

Keep building on what the team has learned by regularly reflecting on what is working, improving existing uses and creating space to test new ideas.

Revisit this regularly, or as part of existing planning cycles, by:

• reviewing recent AI use to identify what delivered value and what didn’t
• improving or extending existing use cases before creating new ones
• identify one or 2 new ideas to test based on team feedback, changing work needs or emerging tools
• updating your AI policy based on what the team has learned in practice
• checking team habits still align with the essential AI practices.