UK government trial of MS365 Copilot finds no clear productivity boost

Ouch!

[Brett_Tabke]

[theregister.com...]

A UK government department's three-month trial of Microsoft's M365 Copilot has revealed no discernible gain in productivity – speeding up some tasks yet making others slower due to lower quality outputs.

[graeme_p]

No surprises there. "AI" can do some simple things fairly well, but not anything complex or that requires actual intelligence.

Add to that the badly thought trial and the British government's less than brilliant IT track record (not all bad, gov.uk is excellent, but most big projects fail or hugely overrun) and I am surprised it worked even this well.

[tangor]

Copilot is well named. After all, the plane is flown by THE PILOT.

As for gubermint work, bureaucracy (and thinking like one) is not brilliant. Rarely comes in on time, always costs more, and hardly qualifies as good.

[thecoalman]

I wouldn't trust any results from a bureaucracy because efficiency does not expand the bureaucracy. You might as well ask buggy whip makers to test out this new thing called automobiles, failure is guaranteed.

[Whitey]

The Register’s report doesn’t surprise.

AI’s ceiling right now: It does the “simple but time-consuming” jobs well (emails, summaries), but anything that needs domain nuance or deeper reasoning is still on us.

Trial design matters: Gov IT trials often flop because success metrics aren’t clear and the change management isn’t there. In that environment, you’ll never see the full potential.

Bureaucratic incentives: Efficiency doesn’t grow bureaucracies. Shrinking workloads can even be quietly resisted.

Copilot branding: @tangor’s right. It’s named Copilot, not Pilot. The marketing primes people to expect too much.

Bottom line: The UK trial says more about bureaucracy than it does about the tech. In leaner, competitive environments where efficiency is rewarded, the same AI tools can deliver real gains.

[graeme_p]

@Whitey. Fans of CS Lewis might remember the professor saying "Its all in Plato". These sorts of discussion make me think "Its all in Yes Minister". Lack of success metrics, growing bureaucracies, budget overruns,..... a lot of it was actually based on real incidents (there are footnotes in the book version).