When AI Makes Teams Slower

Many leaders are seeing the same thing. AI is helping individuals work faster, but team performance isn't improving as expected.

AI makes individual work faster. A task that used to take minutes now takes seconds. But when a whole team starts using AI, the pace of work doesn't always improve. Sometimes it slows down.

That's not because AI isn’t working. It's because making one person faster doesn't automatically make a team faster. Many organisations focus on individual productivity, but the bigger challenge is improving team productivity.

What the research found

In 2025, researchers at METR asked experienced software developers to complete real tasks using AI. They expected AI to save them time. Instead, the tasks took 19% longer.

At the organisation level, Atlassian found a similar pattern. While 89% of executives believed AI had improved productivity, only 6% could point to a measurable improvement across the organisation.

The message is simple: AI helps individuals work faster. That doesn't always mean the business moves faster.

The bottleneck just moves

Imagine one of your team members finishes a report twice as fast. It still needs to be reviewed. It still needs to be approved. Someone still needs to decide what happens next. The faster the work arrives, the more those next steps become the new source of delay. Software teams see this every day. AI can generate code in seconds, but the code still needs to be tested, reviewed, and approved before it reaches customers.

The bottleneck hasn't disappeared. It has simply moved.

(Image generated by AI)

More output isn't always more value

The same thing happens in HR, finance, marketing, operations, and almost every other function. AI helps people produce more reports, presentations, proposals, and ideas. But if decisions don't happen any faster, teams simply end up with more work waiting in queue. More work doesn't automatically create more value.

Focus on where work slows down

Organisations getting the most value from AI aren't just investing in new tools. They're also improving how work moves between people. They reduce unnecessary approvals, clarify decision-making, and help managers remove delays that AI can't solve on its own.

Before introducing another AI tool, ask one question:

Where does work actually slow down?

If the answer isn't at the first draft, another AI tool may not solve the problem.

At ALVIGOR, we help organisations build AI-ready teams by equipping leaders, improving the way work flows, and turning AI into better business performance, not just faster individual productivity.