Fundamentals of good AI culture

A checklist of five things we think every organization needs so staff use AI well, with a question for each to show you where to start.

GuideStrategy & planningGetting startedUpdated October 26, 2026

What an AI culture is

An organization’s AI culture is how its people treat AI in their everyday work. It shows in whether staff feel free to try AI, whether they talk about what they find, and whether they see their leaders using it too. Good tools help, but we think culture does more to decide how much an organization gets out of AI.

Below are five fundamentals we think every organization needs. Each one comes with a question to ask about your own organization. If the answer is no, start there.

This checklist assumes your staff already have access to a capable AI tool on a paid plan. If they don’t, start with how to pick the right AI tools for your team. For a wider framework that also covers tools and workflows, see the six pillars of AI transformation.

1. A clear AI policy

Check: Could every member of staff tell you what they can and can’t do with AI at work?

Writing your rules down removes the guesswork for staff. A clear policy tells them which AI tools are approved, what information must never go into those tools, and what they are free to try without asking anyone. A good policy also takes away staff’s fear of using AI and gives them the confidence to use it in their work.

A policy should also leave plenty of room to experiment. If staff finish reading it and their main takeaway is that they should check with a manager before trying anything, the policy is too restrictive. Write it before you run any training, so people know what is allowed from the start.

Our AI policy worksheet walks through each section of a policy, and the AI policy builder gives you a starter outline to finish with your team.

2. One or more AI champions

Check: Is there at least one named person whose role includes helping colleagues use AI, with time set aside for it?

An AI champion is a staff member whose role includes helping colleagues use AI well. They answer questions, help teams find good uses for AI, and share what works across the organization. Bigger organizations usually need more than one, for example one in each team or office.

A champion supports everyone else’s experimenting. If staff wait for the champion to tell them how to use AI, the organization learns slowly, and most of what it knows leaves when that person does. Give the role dedicated time, too. A champion role added on top of a full job usually gets dropped as soon as work gets busy.

Our guide to why your organization needs an AI champion covers the three roles a champion can play, who to pick, and how to make the role work alongside another job.

3. A place to share AI experiments

Check: Is there somewhere staff regularly share what they tried with AI, including what didn’t work?

Set up a channel where people can post the useful things they did with AI, the things they tried that didn’t work, and what they learned. A Slack or Google Chat channel works well for this.

Ask for failures as well as successes. A post saying an attempt didn’t work saves colleagues from repeating it, and it shows that trying something and having it fail is a normal part of using AI.

Make space in meetings too, for example a few minutes in a regular team meeting where one person shows something they tried. Managers can also ask “What did you try with AI this week?” in one-on-ones.

4. Senior leaders who visibly use AI

Check: Have your senior leaders told staff directly that they want them to use AI, and do staff see leaders using it themselves?

Staff take their cue from leaders. The clearest signal is a direct message from the executive director or CEO saying that staff are encouraged to use AI and that the organization is investing in it. Lead with what AI can do for staff and for animals. Telling staff that AI “will be required in the future” tends to stir up resistance before people have tried it.

Leaders then need to use AI themselves and be open about it. They can share something they built with AI, or use it live in a meeting, including when it gets something wrong.

Research on companies matches what we have seen in our advisory work. In a Prosci study of 1,107 participants, executive behavior was one of the strongest predictors of whether AI initiatives stall or scale. BCG’s 2025 survey of more than 10,600 employees found that the share who feel positive about generative AI rises from 15% to 55% with strong leadership support.

5. Time to learn, booked in the calendar

Check: Does everyone on staff have regular time in their calendar to learn and try AI?

Staff at animal protection organizations are often stretched thin, so learning AI rarely happens unless time is set aside for it. That time can take a few forms:

  • Regular team trainings, where someone teaches the team a tool or technique.
  • Regular team playarounds, where a team spends a set period of time playing around with an AI tool or feature, with no defined goal.
  • Hackathons, events lasting from four hours to several days, where staff work together to solve a particular problem for the organization using AI.

If you’re unsure what to cover, our guide to the 20% of AI features that gets you 80% of the value lists nine features to learn first, in the order we suggest learning them.

You don’t need much to start. A dedicated hour each month, or a four-hour hackathon each quarter, is enough to get going. The same BCG survey found that regular AI use is sharply higher among employees who get at least five hours of training.

Keep going

Related Resources

Want helpputting this to work?

Vegan Hacktivists offers free AI and automation support to animal protection organizations — from a first conversation to a full build.

A beaver carrying a branch through the water