The real path to enterprise AI adoption

Ask many enterprise leaders if their people use AI and you’ll get a confident yes. Ask how many are using it well – you’ll likely get a pause.

Ramp – the fintech company – publishes internal AI fluency levels. L0 is disengaged or performative. L1 is a competent user. L2 is a non-technical builder. L3 is a technical-grade builder.

In 2025, a quarter of their people were L0. Half were L1. Their 2026 target moves everyone out of L0 – staying there is now grounds for dismissal – and shifts the centre of gravity from using AI to it being central in most tasks, every day.

The vast majority of organisations don’t have anything like this. They have a ChatGPT licence and an assumption.

The assumption is that access equals adoption. It doesn’t. Nicholas Bloom’s recent research found that 70% of firms claim to use AI while 80% report zero productivity gains. Senior executives average 1.5 hours a week with the tools. That’s not adoption.

What Ramp understands – and many other don’t – is that fluency is a managed progression, not a natural drift.

You remove friction by giving people tool access without procurement hoops. But to drive adoption you need to make AI a focal point of the business – making it constantly visible through email, Teams and Slack. Creating champions whose job is to evangelise and support. Tracking usage. Discussing it in every team meeting and 1-2-1. Testing for it in hiring.

The good news? Its very possible to go from L0 to L3 in six months. And its a journey every enterprise can make together.

If you’re interested in making that move, get in touch. And we’ll start to chart the course.