AI-generated status reports are everywhere. Here is what they miss.
Most project and collaboration tools now offer some version of "summarise this project" or "write my weekly update". Copilot, Gemini, Notion AI, Slack's AI features and the AI built into project management platforms can read tickets, messages and documents and produce a tidy paragraph about progress. It saves real time, and many people now send updates they did not write.
There is a problem, though. Read enough AI-generated status reports and a pattern appears: they are fluent, positive and oddly empty.
What AI summaries do well
- Collecting activity from many sources
- Writing clear, grammatical prose fast
- Listing what changed recently
For a manager who needs a quick sense of activity, that has value.
What they tend to miss
What is stuck. A summary of activity describes what moved. The most important thing in a status report is what did not move, and why. Items that have been waiting for two weeks produce no activity, so they often vanish from summaries entirely.
Who has the ball. "Design review in progress" hides the real story: the design is done, and it has been waiting on the client for nine days.
Judgement. A person writing a status report decides what matters. A summary weighs everything equally, or by volume of discussion, which rewards noise.
Bad news. Models trained to be helpful tend to write optimistically. "Progressing well with some open items" can describe a project in serious trouble.
Activity is not progress
This is the deeper issue. AI status reports measure activity because activity is what tools can see: comments, edits, messages. Progress is about outcomes and blockers, which often produce no activity at all. A project with lots of chatter and no movement looks healthy in a summary.
Researchers and commentators have also raised the problem of "workslop": AI-generated work that looks polished but shifts the effort of understanding onto the reader. Status reports are a prime example.
A better status report, with or without AI
A useful update answers three questions:
- What finished?
- What is next, and who owns it?
- What is waiting, on whom, and for how long?
The third question is the one that matters most and the one AI summaries handle worst. Answer it yourself, or make sure your tools track it explicitly.
Where a board helps
If waiting is tracked as its own state, any summary, human or AI, can report it accurately. In Axtio, the Other column holds everything waiting on someone else, with the number of days it has been waiting and a flag after a week. A status update written from that board answers question three by design. See visibility vs detailed reporting and how to run async standups.
Use AI, but add the hard part
Let AI draft the summary of what moved. Then add, in your own words, the two or three things that are stuck and who needs to act. That is the part your readers actually need. See also AI productivity tools are overrated.