BMET 27A01 - AI Data and Democracy in North America and Europe

Across Canada, the United States, and Europe, the same words reappear: trust, transparency, ethics. We will ask what they actually do – and whom they exclude.

This is not a policy course. It is a course about the stories that live inside policies, about how documents dream of neutrality, and how code repeats that dream. We will examine AI systems not as tools but as fictions of order. Regulation drafts will be read like screenplays, datasets like novels, and algorithms like political theatre.

Each week moves between institutions and artists: Laurie Anderson’s memory machines, Adam Curtis’s documentary montages, and my own experience working with AI archives and bureaucratic infrastructures. Students are invited to write, design, or code their way through these questions.

Expect a seminar that is half theory, half creative lab – sometimes messy, often alive.

Academic expectations

Come prepared, curious, and willing to experiment. Participation matters more than performance. Each week, students bring a short “trace”: a quote, an image, a dataset, or a brief observation. The course includes one concise oral presentation (10 minutes) and a final mini-audit – a creative and critical analysis of an AI system, interface, or document. No technical background is required.

Grades reward precision and originality rather than technical jargon.