Introducing the Academic Misconduct Index
The AMI is an independent index measuring the prevalence of academic cheating and the quality of institutional responses across 28 countries — the first of its kind.
The Academic Misconduct Index (AMI) launches today as the first systematic attempt to measure academic cheating prevalence and institutional response quality across 28 countries simultaneously.
Why this index exists
Academic misconduct — contract cheating, AI-generated submissions, plagiarism, exam impersonation, collusion, and data fabrication — costs universities, employers, and society in ways that are poorly quantified. Degree credentials lose value when they cannot be trusted. Research findings built on fabricated data mislead subsequent scientists. Professionals enter employment having gained qualifications through fraud.
Despite the scale of the problem, no systematic cross-country index existed. The Corruption Perceptions Index (CPI), published annually by Transparency International since 1995, demonstrated that a well-constructed index can shift policy, generate media attention, and create accountability pressure even without legal enforcement power. The AMI applies the same principle to academic integrity.
What version 1.3 measures
The AMI scores 28 countries on two axes:
- Prevalence Score (P): estimated rate of academic misconduct, built from six dimensions
- Response Quality Score (R): how robustly institutions detect, investigate, and deter misconduct
The six dimensions are: contract cheating (D1), AI-generated submissions (D2), exam impersonation (D3), plagiarism (D4), collusion (D5), and data fabrication (D6).
Key findings from version 1.3
China scores highest on the Prevalence axis (P=100), anchoring the maximum. The UK and Australia score highest on the Response axis (R=88), reflecting their legislative frameworks banning essay mills and mandatory disclosure requirements.
The United States presents an interesting profile: P=52.5, R=51.2, placing it in Q2 ("Aware and fighting it") — the only major Anglophone country in this quadrant, reflecting high prevalence combined with growing institutional response.
Sweden's appearance in Q3 ("Crisis zone") at P=53 may surprise — it is driven by Google Trends data showing unusually high search interest in AI submission tools relative to the country's student population, not by reported misconduct cases.
What comes next
Version 2.0 will add expert perception surveys — modelled on the CPI's eventual addition of expert panels — once this index has an established audience from which credible respondents can be drawn. Coverage will expand beyond 28 countries. Data quality flags will improve as more live institutional disclosure data becomes available.
The methodology document, full dataset, and source code are all available for download and reuse under CC BY 4.0.
Written by Francisco Booth, independent researcher.