Proof, Not Anecdote: Listing AI Skills on Your CV Raises Interview Odds by 8–15 Percentage Points, Oxford Experiment Finds
Proof, Not Anecdote: Listing AI Skills on Your CV Raises Interview Odds by 8–15 Percentage Points, Oxford Experiment Finds
Adding AI competencies to your CV isn't a nice-to-have anymore — it is a statistically significant, causally verified advantage. A controlled hiring experiment from the Oxford Internet Institute finds that candidates who list AI skills receive interview invitations at rates 8 to 15 percentage points higher than identical candidates without them. That effect is large enough to fully offset the hiring penalties typically associated with older age or lack of an advanced degree.
The study lands at a moment when the market signal is already deafening. PwC's 2026 Global AI Jobs Barometer, covering more than one billion job ads across 27 countries, reports that AI-skilled workers now command a 62 percent wage premium — up from 57 percent a year earlier — and that AI-specific roles are growing 69 percent annually, roughly eight times the overall market rate of nine percent. Gloat's workforce data shows the pipeline feeding that demand: the number of workers in occupations that explicitly require AI fluency surged from roughly one million in 2023 to around seven million by 2025.
What the Oxford experiment adds is something the market data alone cannot provide: causal proof that listing these skills changes recruiter behaviour.
The Experiment: 1,725 Recruiters, Three Countries, One Clear Signal
Researchers at the Oxford Internet Institute — led by Fabian Stephany — recruited 1,725 active hiring professionals in the United Kingdom, the United States, and Germany. The study used a paired conjoint design, a method in which each recruiter evaluates two synthetically generated CVs that are identical on every dimension except the attribute being tested. This approach isolates the causal effect of a single variable — in this case, whether the candidate lists AI skills — by randomising all other resume features.
Recruiters were asked a simple question: would you invite this candidate for an interview?
The answer, across occupations and countries, was consistent. Candidates whose CVs included AI competencies saw their interview invitation probability increase by 8 to 15 percentage points compared with otherwise identical candidates who did not list AI skills.
To put that in context: an 8-to-15-point swing in invitation probability is comparable to, and in some cases larger than, the gap researchers typically observe between candidates of different education levels or age brackets.
Important caveat: This is a working paper (submitted to arXiv in January 2026, revised March 2026) and has not yet undergone formal peer review. The conjoint design — while well-established in social science — measures stated hiring preferences via hypothetical resume evaluations, not actual hiring decisions. Real-world recruiter behaviour may differ from survey responses, and the findings should be treated as strong directional evidence rather than settled fact.
That said, the experimental design is rigorous. Randomised conjoint studies are widely accepted for isolating causal effects, and the sample of 1,725 practising recruiters across three major labour markets is substantial.
Not All Roles Are Equal: Where AI Skills Matter Most
The experiment tested three occupations, revealing a clear hierarchy in how much AI skills influence recruiter decisions.
| Role |
Effect Strength |
Recruiter Logic |
| Office Assistant |
Strongest (top of the 8–15pp range) |
Recruiters see immediate, concrete productivity gains from AI fluency in administrative work — document automation, scheduling, data management |
| Software Engineer |
Moderate-to-strong |
AI competency is increasingly expected; listing it confirms the candidate is current, though the baseline assumption of technical ability is already high |
| Graphic Designer |
Weakest (bottom of the 8–15pp range) |
Recruiters express scepticism about AI in creative roles; concern that AI reliance may signal less original creative ability |
The pattern is intuitive. In roles where AI demonstrably automates routine tasks, recruiters reward the signal. In creative roles, the signal is more ambiguous — and some recruiters view it with suspicion.
Certificates Are Nice, but the Label Itself Does the Work
One of the study's more surprising findings: formal AI credentials — whether from a university course or an industry certification programme — produced only a modest additional increase in invitation probability compared with candidates who simply declared AI skills on their CV without formal backing.
In practical terms, a candidate who writes "Proficient in AI-assisted data analysis and prompt engineering" gets nearly as much of an interview boost as one who lists a formal certificate in the same area.
This does not mean credentials are worthless. Formal certificates provided a "significant additional compensatory role" specifically for office assistant candidates, where recruiters may need more external validation of technical claims. But across the full sample, the core finding holds: the presence of AI skills on the CV matters more than how those skills were acquired.
Who Benefits Most: Older Workers and Those Without Advanced Degrees
Perhaps the study's most consequential finding for HR leaders is this: AI skills "partially or fully offset disadvantages related to age and lower education."
The effect was strongest among older office assistant candidates without advanced degrees. For this group, adding AI competencies to the CV closed much of the hiring gap that age and education typically create.
This is a single data point from one study, and it should not be overstated. But the direction is clear: in a labour market that has historically penalised older workers and those without prestigious credentials, AI fluency acts as an equaliser. It signals adaptability and current relevance in a way that traditional qualifications alone may not.
A related moderating factor: recruiters who personally use AI tools in their own work are significantly more likely to favour AI-skilled candidates. As AI adoption spreads among hiring professionals themselves, the premium on candidates' AI skills is likely to grow.
Market Context: The Premium Is Already Priced In
The Oxford experiment doesn't exist in a vacuum. Three independent data sets confirm that the market is already pricing AI skills at a significant premium.
PwC's 2026 Global AI Jobs Barometer (published June 15, 2026) analysed more than one billion job advertisements across 27 countries and found:
- Workers with AI skills earn a 62 percent wage premium over peers without them — up from 57 percent the prior year
- AI-specific job postings are growing at 69 percent annually, roughly 8× the overall market growth rate of 9 percent
- AI-exposed entry-level roles are 7× more likely to require traditionally senior-level skills such as judgment, leadership, and creativity
- These "seniorised" entry-level positions have grown 35 percent since 2019, while other entry-level roles declined 10 percent
Gloat's 2026 AI Skills Demand analysis tracks the supply side:
- Workers in occupations requiring AI fluency grew from approximately 1 million (2023) to 7 million (2025) — a sevenfold increase in two years
- AI skills now appear in 2.5 percent of all U.S. job postings, a 297 percent increase over the decade and growing 20× faster than the overall job market
- 75 percent of knowledge workers already use generative AI tools, frequently without formal company deployment
Together with the Oxford causal evidence, the picture is stark: the labour market is not only demanding AI skills — it is materially rewarding candidates who signal them, even at the resume-screening stage.
Practical Implications
For Job Seekers: What to List and How
List AI skills explicitly. The experiment shows the label matters. Generic technology skills ("proficient in Microsoft Office") do not trigger the same response. Be specific: name the tools (ChatGPT, Claude, Copilot, Midjourney) and the workflows (AI-assisted research, prompt engineering, automated reporting).
Don't wait for a certificate. Self-declared AI competencies produced nearly the same interview boost as formal credentials. If you use AI tools productively in your work, say so — you do not need a course completion badge to claim the skill.
Tailor to the role. The boost is strongest in administrative and operational roles where AI productivity gains are obvious. For creative positions, frame AI as a complement to your original work, not a replacement for it.
If you're an older worker or lack an advanced degree, this matters even more. The data shows AI skills can partially or fully offset age and education disadvantages in recruiter evaluations. It is one of the most effective signals of current relevance available to you.
For HR and Talent Acquisition Teams
Audit your job descriptions. If your roles benefit from AI fluency, say so explicitly. The PwC data shows AI-exposed roles are already being "seniorised" — requiring judgment and leadership skills alongside technical ones. Make sure your JDs reflect this shift.
Recalibrate screening criteria. The Oxford data suggests that AI skills are a stronger signal of candidate quality than many traditional credentials. Consider whether your screening rubrics appropriately weight AI competencies.
Watch for creative-role bias. Recruiters in the study were sceptical of AI skills in graphic design roles. If your organisation values AI fluency in creative teams, ensure screeners are aligned — otherwise, you may inadvertently filter out candidates who have exactly the skills you need.
Train your recruiters. The study found that recruiters who personally use AI are more likely to value AI skills in candidates. Upskilling your TA team on AI tools doesn't just improve their productivity — it changes how they evaluate talent.
FAQs
What is a conjoint design, and why does it matter here?
A conjoint design is an experimental method where respondents evaluate options (in this case, CVs) that vary systematically across attributes. Because attributes are randomised, any difference in outcomes can be attributed causally to the attribute being tested — not to confounding factors. This is what separates the Oxford study from surveys or correlational analyses that show AI-skilled workers do better but cannot prove why.
Do I need a formal AI certification to benefit?
No. The study found that self-declared AI skills produced nearly the same interview boost as formal credentials. Formal certificates offered a modest additional advantage in some roles (particularly office assistant positions), but the primary effect comes from listing AI skills at all — regardless of how they were acquired.
Which roles benefit most from listing AI skills?
Office assistant and administrative roles saw the strongest effect. Software engineering roles saw moderate-to-strong effects. Graphic design and other creative roles saw the weakest effect, likely because some recruiters are sceptical about AI's role in creative work. However, even in the weakest category, the effect was still positive.
What should HR teams change based on this research?
Three things: (1) Update job descriptions to explicitly mention AI fluency where relevant, so candidates know to signal these skills. (2) Review screening rubrics to ensure AI competencies are weighted appropriately — they may be a stronger signal than traditional qualifications. (3) Upskill your own recruiting team on AI tools, since recruiters who use AI personally are more likely to recognise and value AI skills in candidates.
Is this study definitive?
Not yet. It is a working paper that has not undergone formal peer review. The conjoint design measures stated preferences (would you invite this candidate?) rather than actual hiring decisions. However, the methodology is rigorous, the sample is large (1,725 recruiters across three countries), and the results are consistent with broader market data from PwC and Gloat. Treat it as strong directional evidence.
Sources: Stephany et al., "AI Skills Improve Job Prospects: Causal Evidence from a Hiring Experiment," Oxford Internet Institute working paper (arXiv:2601.13286, revised March 2026); PwC 2026 Global AI Jobs Barometer (published June 15, 2026; 1 billion+ job ads, 27 countries); Gloat, "AI Skills Demand in the U.S. Job Market 2026"; iMocha, "Top 25 AI Recruitment Statistics Shaping Hiring in 2026."