Evidence date 30 August 2026
Sources48 cited sources, 50-entry register
All grade A
The Application as an Auditable Claim
What 57 current elite-selection signals reveal about ownership, judgement and authenticity in the age of AI.
Headline figures
The application is best understood as an auditable claim. It opens the inquiry; later stages test whether its facts, reasoning and voice survive scrutiny. Fifty of 57 observations were coded as existing processes at the review date. This classification does not establish when those processes began or whether AI caused changes to them.
Global Elite Selection Monitor: 57 signals drawn from 48 official sources, held in a 50-entry source register, across 23 institutions or systems. Sources checked 30 August 2026. This report uses public institutional material only. It contains no private client evidence and implies no affiliation or endorsement.
Selection is becoming a chain of verification
- 01Selection includes assessment the candidate does not write at leisure.Verified examples include a timed cognitive assessment at KPMG, an assessment centre with a written and a group exercise at PwC Middle East, a case discussion at Strategy& Middle East, essays written on the interview day at the Indian School of Business, an Admissions Written Test before interview at NUS Law, and an in-person Case Challenge at DBS. No proportion is claimed: the inclusion rule and row-level review are not complete.
- 02A change of medium can make a written claim easier to test.A claim written at leisure becomes harder to imitate when it must be explained, adapted, calculated, debated or defended under time pressure.
- 03The AI rules observed concern ownership, not blanket prohibition.Six observations were coded newly explicit about AI. They come from three institutions: DLA Piper, UBS and Cambridge, which permit forms of preparation while protecting independent assessment and truthful representation.
- 04Process design creates unequal translation burdens.Thirteen signals map stage architecture or geographic variation. Candidates must transfer the same underlying evidence across different formats, sequences and markets.
Strong enough for pattern detection, not prevalence
Testing beyond prose, in the institutions' own words
Each statement is confined to what that institution itself publishes.
Nine institutions, nine instruments. Clifford Chance: tests, prerecorded video, assessment day. UBS: online assessments, a pre-recorded video interview, then a final round of virtual or in-person interviews. Deloitte: work simulation and final-stage discussion. Cambridge: interviews exploring thinking beyond the file. Slaughter and May: discussion of a current-affairs article supplied on the day. A&O Shearman: partner case-study interview. Morgan Stanley: how candidates work through a scenario. Yale: evaluative conversation alongside the file.
This does not show that live assessment is new, increasing or superior. It shows that current elite-selection systems already contain substantial verification architecture that becomes more strategically important when polished output is abundant.
This does not establish that live assessment is new, increasing or superior. Fifty of the 57 observations were coded as existing processes at the review date, which records what the page said, not when the practice began.
This does not establish prevalence. The count describes these 57 observations and nothing wider.
Every newly explicit signal defines an AI boundary
| Institution | Permitted or accepted | Boundary |
|---|---|---|
| DLA Piper | Research, structure, personal editing and practice where use is transparent and fact-checked | No completing assessments, replacing thinking, misrepresenting experience or reading scripted answers |
| UBS | Preparation of CVs and cover letters where content genuinely reflects the applicant | Assessments and interviews must be independent; AI answers, recording, transcription, avatars and deepfakes are prohibited |
| Cambridge | Prompts for ideas or assistance with research | Over-reliance risks false claims treated as fraud and weak output that harms the application |
The emerging norm in this sample is bounded assistance plus personal accountability. This is a governance model, not a technology ban.
Processes coded as already existing
50 of 57 observationsobservations coded as existing processes at the review dateTests, cases, interviews and panels appear in the published processes reviewed. No Report 01 observation records when any of them was introduced.
No introduction dates are recorded
No Report 01 observation carries a date of introduction, so no sequence can be established for those 57. Report 02 does contain dated changes: NUS dates its Law test-before-interview rule to 2026 and its Dentistry sequencing to 2021.
Three institutions publish explicit AI boundaries
6 of 6observations coded newly explicit about AI, from three institutionsThis does not establish that AI caused institutions to introduce live, interactive or whole-person assessment. This review does not establish when the Report 01 processes began or whether AI caused changes to them. A small number of Report 02 observations do carry dates, and they are stated where they appear.
The three-gate model turns the research into action
| Gate | Question |
|---|---|
| Admission | Does the written claim clear the threshold and open the inquiry? |
| Progression | Does the claim survive live testing in a different medium? |
| Conversion | Does judgement hold under final-stage scrutiny? |
If AI makes polished output abundant, which forms of assessment reveal judgement - and which merely add cost, complexity and new advantages for the already well prepared?
Methodology and limits
Open the register
Unit of analysis
One row records one verifiable selection signal: a requirement, stage, assessment, institutional statement or material change. Multiple signals may come from the same source.
Source standard
All 48 cited sources in this report are official employer or university pages checked on 30 August 2026. The source register holds 50 entries; two graded entries are retained but cited by no signal. Each observation is narrowly worded to match the source. The live Monitor retains source IDs, URLs, confidence scores and review dates.
Coding used in this report
Report 01 coded signals to categories including live judgement, simulation, case or work simulation, multimodal or technical assessment, problem solving, gamified assessment or whole-person evaluation. The 6/6 measure covers all observations coded as newly explicit.
This is a purposive, cross-sectional sample. It overweights the UK, global careers pages, law and banking. Published process pages may omit informal practices and may describe different programmes that are not directly comparable. The dataset cannot establish prevalence, causal change over time, candidate outcomes or institutional intent beyond the stated process.
Publication rule
Market-wide claims require triangulation across institutions and sectors. Historical trend claims require longitudinal evidence. This report therefore distinguishes what the sample shows from what remains a hypothesis.
Akram, H. (2026). Report 01: The Application as an Auditable Claim. Who Gets In?, Elite Careers Strategy, Global Elite Selection Monitor. www.accessecs.com/who-gets-in/report-01/ Evidence checked 30 August 2026; revised 31 August 2026.
Source register SRC-001 to SRC-048 cited; 50-entry register, all checked on 30 August 2026 and graded A - official primary. Revised 31 August 2026. No client material is included. Findings remain subject to the stated methodology and limits.