Make the whole company AI-mature

In most companies AI adoption has already started, unevenly. A few people are getting real value from it, a few teams have quietly banned it, and the rest are waiting for someone to tell them what is allowed and what is worth their time. Leadership has bought licences and is wondering why productivity has not moved.

The gap is not the technology. It is that nobody owns the change: which tools, which rules, who trains whom, and how you know it is working.

• Agree the guidelines. With business stakeholders, security, privacy and leadership: what people may use AI for, with which data, and where a human has to check. Written so people can actually follow them.
• Choose the tooling and the settings. Reviewed together with security and privacy: which tools to connect, which integrations to switch on and which to leave off, so the rollout is safe and still genuinely useful.
• Manage the contract and the cost. Negotiate with the vendor, size the licences to real use, and keep the spend in hand as adoption grows.
• Build the training programme with the people organisation. Short, role-specific sessions: what it is good at, what it is bad at, and how to check its output.
• Sit down with individual users. AI skills labs and one-to-one sessions where people bring the task they are stuck on and leave with it done differently. This is where adoption stops being a mandate and starts being a habit.
• Build the internal champions. A network of people in each team who carry it on after I step back.

I rolled out Claude across a company to hundreds of employees. The guidelines were worked out with business stakeholders, security, privacy and leadership, and the tooling and its settings were reviewed with security and privacy before anything was switched on. I negotiated the contract, which included a substantial discount for the client, and kept the cost in hand as usage grew. With the people organisation I set up the training programme, and I sat down with many individual users to help them succeed with AI on their own use case. Rolling out at scale with a personal, tailor-made approach is what made the difference: usage spread beyond the early adopters and into daily work across the organisation.

• An AI assistant in use across the company, with adoption you can measure and productivity to show for it.
• Guidelines that security, privacy and leadership have signed off, and that people follow because they understand them.
• A contract and licence count that fit how the company actually uses it.
• Teams that know what to use it for and how to check the result.
• Internal champions and a training format you can keep running.