Computational Methods For Public-Interest Research
What Platforms Hide
How can independent researchers study platforms they cannot fully inspect?
About the talk
My new book The App Economy (Bristol University Press, 2026, open access) asks how Apple and Google use their dominance over the app economy for their bottom line, with what consequences for society, and how to respond. In this talk, I zoom in on tracking and platform power. Drawing on the book, my TrackerControl app and ongoing research, I explore who controls our apps and data – and how technical tools, law and independent research can help us challenge that control.
About the talk
Political creators may be paid to promote candidates or earn money through livestream gifts, yet these arrangements can be difficult to detect. The proposed Election Monetisation Observatory (EMO) would bring together open-source tools to study hidden advertising in election content, drawing on the expertise of researchers, creators and journalists. This talk considers what those tools could reveal, and where limited platform data and changing features still get in the way.
About the talk
Consumers increasingly use AI chatbots for advice on what to buy. With companies like OpenAI and Google monetising their AI through advertising, this raises difficult questions about the bias and impartiality of such advice. In response, we conduct an AI audit of popular chatbots using real commercial-advice queries.
About the talk
AI labels are emerging as a primary safeguard for transparency about AI-generated content on social media, including under the EU DSA and AI Act. Yet, limited systematic evidence exists on how platforms implement such labels in practice. In response, we conducted a large-scale legally grounded audit of AI labelling practices across several major social media platforms. In this talk, we will share insights on the different methods used in this study, and the types of signals that are currently (in)visible to external auditors.
Why this matters
App stores set the terms for mobile apps. Social media platforms influence which political posts people see. Chatbots offer shopping advice.
These platforms shape everyday choices and public debate, but control much of the data needed to examine them. Researchers can measure app traffic, inspect public posts and audit chatbot answers, yet each gives only a partial view.
This is platform observability in practice: what can we see and test from the outside, and what can those observations actually establish? The talks bring technical, legal and social perspectives to that question, including where current methods fall short and what might make the next study possible.
Group exercise
What should we study next?
Bring a question from your work, or start with one from the talks. In small groups, we will ask what we can observe, what better access would change, and how methods from different fields might help.
An app sharing data, a monetised political video, an unlabelled AI post or a chatbot's product suggestion.
App traffic, public posts, AI labels or chatbot answers: which would help answer your question?
AI and NLP, network analysis, law and social research each bring different tools and questions.
Access, privacy, weak labels and changing platforms. Could audits, shared datasets or collaborations help?
Takeaway
Questions worth testing, possible methods and a clearer sense of the evidence still missing.
Provisional agenda
How the session will run
Audience
For anyone interested in studying platforms with computational methods. No background in law or policy is needed.
Organisers
Interested in joining?
What would you like to discuss?
Planning to attend BNAIC? Share your area of work or a platform question you would like to explore. This interest form is separate from conference registration.