UCL Academic Statement Example: Algorithmic auditing researcher to AI accountability (Score 93)

During my final undergraduate year in Computer Science, I encountered a problem that my technical training alone could not resolve. I was auditing a recidivism-prediction model used in a simulated criminal justice context, mapping feature weights, testing for demographic parity, and documenting disparate impact across protected groups. The audit methodology was sound, but when I tried to translate findings into a recommendation that a policy audience could act on, I had no framework for doing so. That gap between what an algorithm does and what governance structures should require of it is the question I want to pursue at postgraduate level, and it is the reason I am applying to UCL's MSc in Artificial Intelligence Governance. My undergraduate dissertation, developed within my university's algorithm audit research group under faculty supervision, examined how existing audit frameworks map onto regulatory obligations in high-risk AI deployments. Using a mixed-methods approach combining structured literature review with comparative analysis of audit outputs across three published case studies, I produced an accountability memo identifying where technical audit evidence fails to meet the evidentiary standards regulators actually require. The core finding was that most audit reports are written for engineers, not for the oversight bodies that must act on them. That translation problem became the organising question of my subsequent work. A three-month placement with an AI governance advisory team gave me a different vantage point. I prepared stakeholder-facing briefing notes comparing implementation risks across regulatory regimes, and I observed directly how the same audit evidence was interpreted differently depending on whether the reader held a legal, technical, or policy background. The briefing note I produced was used in an internal planning discussion on audit disclosure requirements. Alongside this, an applied project on algorithmic auditing produced a portfolio artefact connecting technical audit methodology to accountability outcomes, which I presented at a peer workshop organised through my university's AI governance student initiative. UCL's MSc in Artificial Intelligence Governance is the right programme for this trajectory for specific reasons. The module on AI regulation and policy directly addresses the evidentiary translation problem I identified in my dissertation, and the research methods component will allow me to develop the mixed-methods audit design I used at undergraduate level into a more rigorous framework. The programme's interdisciplinary structure, drawing on law, computer science, and public policy, matches the multi-audience problem I encountered in practice. I am particularly interested in working within UCL's research environment on questions of audit transparency and regulatory adequacy, areas where the programme's faculty have published work I have already engaged with in my literature review. My goal is to contribute to the design of audit frameworks that are both technically credible and institutionally legible, work that sits at the intersection of computer science and governance. The MSc is the necessary next step: it will give me the regulatory and methodological grounding to move from producing audit evidence to shaping the standards by which that evidence is evaluated.