Your team doesn't want to use AI. Now what?

The CTO buys licenses, sends an email 'we're using AI from now on,' and a month later two people out of twenty use it. The rest tried it once, got bad output, and went back to what they know.

I see this so often I could call it the standard scenario. And the problem isn't the people — it's the approach. Mandating adoption from the top doesn't work. What does?

Why developers resist

Resistance to AI isn't irrational. It has concrete causes that you need to understand before you can address them.

Bad first experience

They tried one prompt ('fix this bug'), got generic code that didn't work in their context, and dismissed the entire tool. This is the most common cause. The problem isn't the tool — it's the prompt. But who teaches them that?

Fear of replacement

'If AI can write code, why would they need me?' This is a real concern that shouldn't be dismissed. The answer: AI replaces routine tasks, not developers. A developer who uses AI is more productive — not more replaceable. But this needs to be said explicitly.

No time to experiment

They have sprint commitments and learning a new tool feels like a luxury. If management doesn't explicitly allocate time for experimentation, people simply won't do it. Sprint commitments always win.

What works: three proven strategies

1. Find enthusiasts and give them space

Every team has 2-3 people who enjoy experimenting. Give them time and space — explicitly, not 'when you have a moment.' Let them find use cases that work on your code. Then let THEM show the rest of the team.

Peer recommendations work 10x better than a manager's email. When Tom says at standup 'yesterday I generated 40 tests with Claude Code in an hour, which would have taken me two days' — that convinces.

2. Show concrete savings, not potential

'AI will save you time' convinces nobody. It's too abstract. What convinces:

• 'Tom generated 40 tests in an hour — manually it would have taken 2 days'
• 'Sarah cut review time from 2 days to 4 hours with AI pre-review'
• 'Martin understood a new module in 30 minutes instead of half a day'
• 'The whole team deployed twice as many features this sprint'

Concrete numbers from YOUR team. Not internet benchmarks, not vendor promises. Your numbers, your people, your code.

3. Don't make it mandatory

Top-down mandates create compliance, not adoption. People will 'use' AI just to hit a metric — but not effectively. Instead, create an environment where it's easy to start and where the results of those who started are visible.

In practice: a shared Slack channel for AI tips. At retrospective, ask 'where did AI help you?' Celebrations when someone finds a great use case. No 'you must use AI' — but 'look at what Tom did with it.'

Realistic adoption timeline

Realistic adoption isn't 'everyone uses it within a week.' It's a gradual process:

• Month 1: 2-3 pioneers experiment, find use cases
• Month 2: they share results, 5-6 more try it
• Month 3: half the team actively uses AI for something
• Month 4-5: AI becomes part of the standard workflow
• Month 6: most of the team uses AI daily

This isn't slow — this is sustainable. Rapid adoption mandated from above leads to superficial usage and a return to old habits.

What about those who 'will never use AI'

Every team has someone who says 'I don't need AI.' And that's OK. Don't force them. But make sure they have access and training when they decide to try. Often what happens is the skeptic sees colleagues saving time and thinks 'let me try it on that boring task' — and stays.

The best adoption isn't forced. It's contagious. When people around you save hours every day, you want to know how.

Mistakes I see repeatedly

• 'We use AI now' email with no training or support
• AI usage metrics as KPIs — people optimize the metric, not productivity
• Expecting immediate results — adoption takes 3-6 months
• Ignoring resistance instead of understanding it — resistance has reasons, address them
• One tool for everyone — different people need different tools

AI adoption in a team is a culture change, not just a tool change. And culture change requires time, patience, and the right approach. You can't mandate it — but you can inspire and enable it.