TL;DR

AI minister Kanishka Narayan has made contingency planning for large-scale AI job losses one of his priorities, while stressing there is no evidence yet of an overall fall in UK employment. Speaking at a Labour conference fringe event, he suggested the effect may be hidden by staff quietly using AI at work. For UK employers, the signal is that hiring decisions shaped by automation are now on the government’s radar, even where headcounts are not falling.

What the minister said

Narayan told the meeting the government must “take seriously the possibility of an unprecedented impact on the jobs market”. His reasoning was about timing: “If it’s going to hit us in a big way, you are going to want to be as ahead of the curve as possible.”

He offered two reasons the official picture may understate things. Industry surveys miss the “silent adoption” of AI by individual workers. And while firms report no aggregate impact, one tech company had told him automation meant it now recruited fewer staff. A contingency plan could reach into “labour market policy” and “even where we might go on some aspects of regulation”, he said, and he promised “much better transparency” when organisations adopt AI, so productivity gains can be measured.

The think tank’s numbers

The IPPR’s Carsten Jung gave the room a sharper range. On his figures, chatbots already put about 11% of UK jobs in the “highly exposed” bracket, with secretarial and admin roles at the front, while agentic AI could lift exposure to “about 60% of tasks across the UK economy”. In a worst case, about 8 million jobs could be hit. Because more women work in service jobs, he said, they are “40% more exposed than men”.

Jung admitted IPPR’s 2024 scenarios had been partly wrong, and said the institute is now writing a “pandemic preparedness” style report on labour disruption. His policy ideas include retraining for service workers and changing tax rules so that keeping staff looks better value than automating them. Mike Clancy of the Prospect union welcomed a bigger say for workers but warned against leaving them as “victims” of the shift.

Looking forward

A contingency plan is a hedge, not a forecast, and the minister was careful to frame it that way. The harder test is measurement: if AI mainly shows up as roles that are never advertised, neither surveys nor redundancy figures will catch it. Firms can expect questions about how AI is changing their hiring, not just their headcount, as that transparency promise takes shape.