The Meeting Problem: How AI Is Changing How Decisions Get Made

A large share of organisational meetings exist to do one thing: get everyone the same information so a decision can be made. People gather to share updates, present analysis, bring each other up to speed, and only then actually decide. AI is quietly dissolving the reason for the first part. When information can be gathered, synthesised and shared without a meeting, the meeting that was built around doing that has lost its purpose - but most organisations are still holding it.

This article looks at how AI is changing how decisions get made, and how to rethink meetings around it.

What meetings were really for

It is worth being honest about what meetings actually do. Some meetings exist to build relationships, align on direction, or have a genuine debate - those have clear value. But a great many exist to move information: status updates, read-outs, briefings, “let me walk you through this.” The decision, when it comes, often takes a small fraction of the time. The rest was logistics - the cost of getting everyone to the same understanding.

AI changes the economics of that logistics.

How AI changes the picture

AI can do much of the information work that meetings were built around. It can synthesise updates and analysis into a shared briefing before anyone meets. It can answer the clarifying questions that used to eat meeting time. It can surface relevant history and context on demand. And it can capture and distribute what was decided without someone writing it up afterwards.

What this leaves is the part AI cannot do: the judgement, the debate, the alignment, the decision itself. If information-moving happens before the meeting, the meeting can be about what only people can do.

The new shape of decision-making

This points to a different rhythm. The information work happens first, asynchronously, with AI help - everyone arrives already informed. The meeting itself is shorter and exists for the human part: surfacing disagreement, weighing trade-offs, applying judgement, deciding. And the record of what was decided and why is captured without a separate effort. Fewer meetings, shorter meetings, and meetings that are actually about deciding rather than about preparing to decide.

Where this goes wrong

There are failure modes to avoid. The first is keeping the old meeting and adding the AI prep on top, so people now do the pre-reading and still sit through the read-out - more work, not less. The second is over-correcting: cutting the meetings that genuinely needed to be meetings - the ones for debate, alignment and relationship - because they got lumped in with the information meetings. The third is assuming that because everyone could be informed, everyone is - AI makes shared understanding possible, but people still have to actually engage with the briefing. The redesign needs judgement, not a blanket rule.

Decisions still need humans

It is worth being clear about a boundary. AI changes how the information for a decision gets assembled and shared. It does not make the decision. The judgement about what the information means, which trade-offs matter, what the organisation should value, and who is accountable for the outcome - that stays human. The goal of rethinking meetings is to spend more human time on exactly that, not less.

What leaders should do

If you are responsible for how your organisation runs, look at your meetings and ask which exist mainly to move information. For those, shift the information work to before the meeting with AI help, and shorten or remove the meeting itself. Protect the meetings that genuinely need to be meetings. Make sure people actually engage with pre-meeting briefings rather than assuming they will. And keep the decision - the judgement and the accountability - firmly with people.

The bottom line

Many meetings exist to move information so a decision can be made, and AI is dissolving the reason for that work to happen in a meeting at all. The new shape of decision-making puts the information work first and asynchronously, and reserves the meeting for the human part - debate, judgement, deciding. Organisations that redesign around this will spend less time meeting and more time deciding well. Those that add AI prep on top of the old meeting will just have more work and the same long meetings.