Microsoft’s AI poll points to a hiring gap at the bottom of the ladder

On 20 October, Microsoft published research for its Frontier Tour in London that set business leaders against their own staff. Some 90% of business leaders expect AI to “impact UK society positively over the next five years”, against around 49% of employees. The work was commissioned by Microsoft, which sells the Copilot AI assistant, and carried out by Censuswide, which polled 5,000 UK employees aged 18 and over and 1,000 senior business decision-makers between 2 and 12 October 2026.

Darren Hardman, who runs Microsoft’s local business as chief executive, is quoted in Microsoft’s write-up, which presents the results as a split in confidence between leaders and staff. On the leaders’ side, around 90% report having a clear AI strategy. Among leaders in organizations already adopting AI, 84% report moderate or significant changes to how work is done, and 36% say the changes have been significant.

Among employees who use AI at work, 71% report a positive impact on their work or career and 69% say AI has improved the quality of their work. The same group is warier about what happens around them: 35% worry the technology has lowered their own job security, and 57% are concerned it will reduce job opportunities for young people and new entrants.

Employees who use AI at work have the most direct experience of the tools. A majority of them (71%) rate it for their own work, and a majority (57%) are also concerned it will reduce openings for young people and new entrants. For hiring, the second figure carries more weight than any gap in confidence between leaders and staff, and a separate employer survey published in October points the same way.

The users’ view of entry-level jobs

The 71%, 69%, 35% and 57% cover only employees who use AI at work, while the published findings do not give a base for the 49%. The most pessimistic finding about young people therefore comes from the group with direct experience of the technology, most of whom report positive results from it, which makes it harder to put down to unfamiliarity or fear of the unknown.

What that group describes is a question of job design and recruitment: which tasks junior staff still do once a tool drafts, summarizes and checks, and whether a business keeps hiring people to learn on those tasks. In a mid-market firm those decisions sit with the managing director and the people director. Microsoft’s survey records expectations, though. It does not show whether junior hiring has fallen at the organizations it polled.

The cost case for experienced hires

Employer data on hiring came out a week before Microsoft’s research. The CBI/Pertemps Employment Trends Survey 2026, published by the CBI on 13 October, found that 44% of businesses say rising recruitment and onboarding costs are pushing them toward experienced hires, which the CBI links to fewer entry-level opportunities for young people. Over half (53%) say wider cost pressures mean they are offering fewer apprenticeships than they would like. The survey drew responses from 682 firms between August and September 2026.

The CBI puts the shift down to cost. Its 44% figure says nothing about AI, and the two surveys asked different people different questions, though both point the same way. Asli Atay, senior policy advisor at the Work Foundation at Lancaster University, told People Management on 20 October that “Organisations risk becoming diamond-shaped, with fewer people at the bottom”.

For a business of a few hundred people the bottom of that diamond may be a handful of graduates, apprentices or junior hires a year. When recruitment costs favor experienced candidates and AI tools take on routine first-year work, that intake can shrink without anyone announcing a decision to cut it.

What the strategy figures leave out

Around 90% of leaders in Microsoft’s poll say they have a clear AI strategy. The published findings do not say how many of those strategies address entry-level hiring, apprenticeships or how junior staff are meant to learn work that software now does in a first draft. For the 84% of leaders in adopting organizations who already report changes to how work is done, the consequences are likely to show up in who they recruit as much as in which tools they license.

The figures most favorable to AI at work, the 71% and 69%, come from research paid for by a company that sells AI software. The 57% finding on young people and new entrants cuts against that commercial interest, and it is the one Microsoft’s own framing as a confidence gap does least to explain.

On the evidence published in October, the staff who use AI and the employers who pay for recruitment point toward the same outcome from opposite ends: fewer ways in for people starting out. Microsoft called it a gap in confidence between leaders and staff. Set beside the CBI’s data, its employee numbers read as an early warning about entry-level roles, and any shortage that follows is likely to surface in mid-market management ranks several years after the junior intake thins.