Research published by Elite Careers Strategy · Revised 19 September 2026
What is changing in early-career selection and work?
Deloitte and PwC's dated official statements document changes in intake, selection and work. They do not establish that AI caused changes in hiring volumes.
This Observatory preserves what institutions publish, the population each statement covers, and the limits of comparisons over time.
Do published intake figures establish AI-driven graduate hiring cuts?
Official Deloitte UK announcements show >1,700 early-career programme joiners in 2023, ~1,400 in 2024, 1,498 announced across September/December 2025, and >1,500 in a broader 2026 early-career population including technical academies. Official PwC UK material reports 1,618 graduate + 175 school-leaver places in 2023, >1,500 graduate/school-college joiners in 2024 and ~1,300 in 2025. Neither firm's cited intake sources establish AI as the cause of the changes.
Limit: Populations and wording differ by year; figures are not pooled into a market statistic.
Does PwC’s evidence establish when interviews changed and why?
PwC's September 2024 release stated that graduates applying across its business areas would have a final-stage in-person interview in the coming recruitment season. Its 2025 release described a skills-enabled redesign of the graduate offering. Current 2026 selection protects independent assessment performance, while the Strategy pathway names technical, AI and human skills.
Limit: The 2024 release establishes the announced policy. Without a comparable earlier interview-policy observation, it does not establish when this stage was first added. The sequence does not prove AI caused the policy or redesign.
What do employers say about AI and early-career work?
PwC's September 2026 financial-results statement says firm-wide AI tools and learning are helping people take on higher-value work earlier in their careers. Deloitte's 2025 intake introduced new joiners to its GenAI platform and its 2026 induction explicitly highlights judgement, creativity, empathy, adaptability and technology change.
Limit: These statements establish employer-described work/capability change, not displacement of junior jobs or proof that human judgement has become more valuable.
Deloitte's 2026 headline above 1,500 includes technical-academy participants. It must not be treated as the same population as every earlier intake headline.
Published places, announced joiners and completed starts are different measures. The PwC series preserves the source wording rather than calculating a pooled percentage decline.
Continued hiring neither disproves contraction nor establishes that AI has no employment effect.
Employers describing AI-enabled work is evidence of their stated practice, not independent proof of productivity gains, job displacement or lost training opportunities.
Deloitte UK: what does the dated evidence show?
Use like-for-like September intake announcements as the primary comparable series. Broader annual-report hiring counts are retained separately because their population/window differs.
early careers including graduate + apprentices + Industrial Placement + technical academies
more than 1,500 good judgement; creativity; empathy; adaptability; technology change; technical academies recruit for potential including people without prior technology experience
1,353 graduates/apprentices/industrial placements + 130 academy participants progressed into technology teams + 20 academy entrants later in month 2026 headline population includes technical academies, so do not compare the >1,500 headline mechanically with the narrower 2023-25 programme populations.
over 1,500 joined; 450 summer interns PwC said all graduates applying across business areas would have a final-stage in-person interview with a partner/director at one of 19 UK offices for the coming recruitment season.
Source frames this around value of in-person experience; it does not attribute the change to generative AI.
around 1,300 joined; 433 summer interns PwC explicitly described itself as becoming a skills-enabled organisation and redesigning its graduate offering to simplify entry, increase choice and match paths to skills/interests, while acknowledging unequal access to networks.
PwC said it simplified pathways into the firm to create a clearer/more accessible candidate experience. PwC said AI tools had been rolled out to every employee with AI learning for staff/partners and that the investments were helping people take on higher-value work earlier in their careers.
Direct employer evidence linking AI deployment to earlier-career work content; does not establish headcount substitution or loss of apprenticeship.
Early Careers assessments must be completed independently; AI/third-party generation, editing, rewriting, summarising or coaching of assessment responses is prohibited unless authorised. PwC states it does not use AI to assess candidate performance in Early Careers selection.
PwC says structured learning builds critical technical, AI and human skills; applicants from all degree disciplines are welcomed for Strategy except specialist Economics Consulting module requirements.
The existing source-monitoring and production-check workflow completed successfully on 19 September 2026. This verifies execution of those steps, not every source observation or an autonomous research-to-publication loop. Source changes still require evidence and scope review. End-to-end continuous research publication remains unverified.
How can you use and cite this research?
Who Gets In? (Elite Careers Strategy), What is changing in early-career selection and work?, revised 19 September 2026. Cite the institution, period and population alongside any figure.
Research programme of Elite Careers Strategy