Research

Measuring harmful content, recommender system effects, and platform interventions at scale. We build methods to hold platforms accountable to the public.

Applying artificial intelligence to detect, model, and mitigate emerging threats in online environments.

Auditing moderation systems and developing fairer, more transparent approaches to detecting and removing harmful speech.

Characterising new risks introduced by generative AI systems — including synthetic media, automated disinformation, and misuse at scale.

Studying how at-risk populations — especially minors — are exposed to harmful content and dark design patterns on social platforms.

Building public datasets, data-donation methods, and infrastructure for independent platform research and regulatory accountability.