The issues I have centered my research and career around – the legitimization of harmful movements and language, information ecosystems, radicalization and mobilization, and democratic backsliding – have increasingly demanded closer analysis of cyberspace. This includes online media algorithms, which can dictate the content that appears on a person’s page. It includes biases from data sets that large language models draw from to generate content. It also includes the role of media in the creation and spread of misinformation and disinformation. My entry point into this work comes from the conviction that the people with backgrounds in policy, democracy, and human rights are essential to ensuring the responsible development and governance of online spaces, from social media platforms to online gaming spaces to artificial intelligence models.
Where My Research Meets AI Systems
Radicalization Pathways and Recommendation Architecture
My academic work has traced how far-right movements build platforms and recruit adherents through specific narrative strategies — the weaponisation of women’s rights rhetoric, the appropriation of democratic language to advance exclusionary agendas, the use of grievance framing to normalise extremist positions. These strategies do not operate in a vacuum. They operate within information environments shaped by algorithmic systems that optimise for engagement, surface content based on predicted relevance, and create feedback loops between user behaviour and content visibility.
Understanding how a recommendation algorithm selects and sequences content is not a separate question from understanding how radicalisation pathways form. It is the same question asked at a different layer. The narrative dynamics I have studied at the political and sociological level — escalation patterns, in-group reinforcement, gateway content that appears moderate but leads to increasingly extreme material — map directly onto observable platform behaviours that emerge from optimisation decisions made during system design. Analysing these dynamics requires both the domain knowledge to recognise what is happening in the content and the mechanistic understanding to explain why the system surfaced it.
Content Moderation and the Limits of Classification
At Outright International, I monitored how anti-SOGI and anti-SRHR constituencies operated within the UN system — deploying language carefully calibrated to appear within the bounds of legitimate political discourse while advancing agendas designed to exclude. This is structurally identical to one of the central challenges in automated content moderation: distinguishing between protected speech and harmful speech when the harmful speech is deliberately constructed to resemble the protected kind.
Automated classification systems — whether keyword-based filters, machine learning classifiers, or large language model evaluators — operate on surface features of language. They can identify slurs, flag known harmful phrases, and detect patterns associated with previously labelled harmful content. What they struggle with is context, intent, and the kind of strategic ambiguity that sophisticated actors exploit. My research on how extremist movements weaponise the language of rights and legitimate concern is directly relevant to understanding why these systems fail at the cases that matter most, and where human judgment, policy design, and system architecture need to compensate for those limitations.
Gendered Harm and AI System Behaviour
My MPhil dissertation examined how women’s rights rhetoric is co-opted within right-wing movements to justify xenophobic narratives — a process in which the language of gender equality is repurposed to serve exclusionary ends. This dynamic has direct parallels in how AI systems handle gendered content. Training data reflects the biases present in the text it was drawn from: patterns of representation, association, and framing that encode gendered assumptions into model behaviour. The outputs that result — from hiring tool recommendations to generated text that reproduces stereotypical associations — are not random failures. They are the predictable consequences of specific data and design choices.
Understanding the mechanism matters because it determines what interventions are possible. A system that produces biased outputs because of training data composition requires a different intervention than one that produces biased outputs because of optimisation objectives or post-training alignment choices. The analytical instinct I bring from gender justice research — asking not just what is happening but how it is being produced and whose interests the framing serves — applies directly to examining AI system behaviour with the specificity needed to recommend actionable changes.
Analytical Approach
-
Mechanism over observation: Documenting that an AI system produced a harmful output is necessary but insufficient. The question that matters for governance and intervention is why — what in the training data, the optimisation objective, the safety architecture, or the deployment context produced that specific output. I approach AI systems the same way I approach political movements: by tracing the causal chain from structural conditions to observable behaviour.
-
Comparative analysis across systems: Just as my academic work compared nationalist movements across the US, UK, and Türkiye to identify both shared patterns and context-specific dynamics, I examine AI systems comparatively. Different models make different design choices about safety, content boundaries, and output behaviour — and those differences reveal the trade-offs and values embedded in each approach. A single system in isolation reveals less than multiple systems responding to the same input.
-
Scepticism toward surface-level solutions: My experience monitoring opposition to human rights initiatives at the UN taught me that the most consequential dynamics often operate beneath the visible layer — in procedural manoeuvres, strategic abstentions, and coalition-building that doesn’t appear in the public record. I bring the same scepticism to AI safety claims: understanding not just what a safety measure is designed to do, but what it actually prevents, what it misses, and what sophisticated actors can route around.
-
Rights frameworks as analytical tools: International human rights law and norms provide a rigorous framework for evaluating AI systems — not as abstract principles, but as specific, operationalisable criteria for assessing whether a system’s behaviour meets standards that the international community has agreed to. My training in applying these frameworks to complex political situations translates directly to evaluating AI governance proposals, safety policies, and deployment decisions.
Current Focus Areas
-
AI governance and regulatory frameworks: Engaging critically with the EU AI Act, emerging UN frameworks on AI and human rights, and national-level regulatory approaches — assessing what each mechanism requires, what it leaves unaddressed, and where the gaps create space for harm. My background in UN advocacy and international law gives me the institutional fluency to understand not just what these instruments say, but how they function in practice.
-
Trust and safety as a systems problem: Examining how content moderation, platform integrity, and responsible AI operate as interconnected systems rather than isolated functions — and how the dynamics I have studied in offline political movements (coordinated inauthentic behaviour, narrative laundering, exploitation of procedural ambiguity) manifest in platform environments.
-
AI and extremism: Investigating the specific mechanisms by which AI systems interact with radicalisation dynamics: how recommendation systems can create or accelerate exposure pathways, how generative AI can be used to produce extremist content at scale, and how the design choices that shape these behaviours can be identified and addressed.
-
Alignment and safety from a human rights perspective: Examining how approaches to AI alignment — the techniques used to shape model behaviour toward intended outcomes — interact with questions of whose values are encoded, who participates in defining acceptable behaviour, and how the process of alignment itself reflects and reproduces existing power structures.
Writing & Analysis
I’m building a body of writing that examines how AI systems interact with political extremism, human rights frameworks, and governance challenges.
Current sample
Roblox Product Policy Memo applies platform policy to coded misogyny, extremist dog whistles, and moderation edge cases on Roblox. It reflects the kind of work I’m interested in doing more of: product-grounded analysis that moves from policy language to concrete enforcement decisions.