AI Ethics & Governance
The hardest questions about AI are almost never technical questions.
A supporting track, and the one furthest from the technical desk. But precisely because the daily work is building models, helping weigh what they end up being used for feels like an obligation.
A technical decision travels outward: through organisations, then policy, then people at large. Whoever bears the consequence usually sits furthest from where the decision was made.
Long before the governance papers came Indonesian-language essays on the post-literate society and the promises of Industry 4.0, written in a critical register, not a promotional one. The unease has not changed since: technology arrives faster than our capacity to agree on what it is for.
The most recent work proposes a society-oriented model of AI governance: instead of one chain of command from regulator to developer, a parallel layered model in which societal values, policy, organisations, and technical systems continuously feed back into one another. The framework appeared in Frontiers in Artificial Intelligence and produced a T20 policy brief on AI adoption in healthcare.
On the empirical side, the subject is what large language models actually do to how students learn, including higher education in the Global South, where assumptions that hold on Western campuses often do not hold at all. Another branch concerns disinformation and hate speech in Indonesian digital space, particularly around elections.
AI governance
A parallel layered model with multi-actor coordination, and its application to AI adoption in healthcare.
LLMs & higher education
Empirical study of student LLM use, with particular attention to the Global South context.
Disinformation & digital space
Detecting hoaxes, hate speech, and electoral disinformation in Indonesian-language social media.
Technology critique & literacy
Essays on the post-literate society, ecoliteracy, and critical readings of the Industry 4.0 narrative.
About supervision on this track
This track runs through collaboration and writing rather than thesis supervision, so there are no open topics here. If this is what draws you, students usually enter via the Sci-ML or Cultural Heritage track first and carry the ethical questions into the project. For research or policy collaboration, email is the way in.
Publications on This Track
A selection from the full list on Google Scholar.
Society-oriented AI governance: a parallel layered model and multi-actor coordination framework
A society-oriented model of AI governance, layered and parallel rather than a single chain of command, which produced a T20 policy brief on AI adoption in healthcare.
Advancing global south university education with large language models
An empirical look at what large language models actually do to how students learn, with attention to Global South higher education, where Western-campus assumptions often do not hold.