Civic Action and AI

Democratic societies have a mechanism for collective decisions about how technology shapes life: civic engagement. Voting, advocacy, public comment on proposed regulations, journalism, community organizing, and litigation are not soft alternatives to the 'real' technical work of AI development — they are levers of power that shape what AI systems get built, under what conditions, and accountable to whom. Understanding these levers and how to use them is part of what it means to be an educated citizen in an AI-shaped society.

The relationship between citizens and AI governance is not one-directional. AI systems shape what citizens see, believe, and decide — through recommendation algorithms, content moderation, news curation, and information retrieval. At the same time, citizens — through their representatives, courts, markets, and collective action — shape what AI systems are permitted to do. Understanding both directions of this relationship is essential.

How Citizens Have Already Shaped AI Policy

Civic action on AI is not theoretical — it has produced concrete outcomes. Bans on facial recognition: Between 2019 and 2023, more than a dozen American cities — including San Francisco, Boston, and New York City — enacted restrictions or bans on government use of facial recognition technology. These outcomes did not emerge from technical debates inside companies. They resulted from advocacy campaigns led by civil liberties organizations, journalists who reported on error rates and civil-rights implications, residents who attended city council meetings and testified, and legislators who responded to constituent pressure. The same technology exists in cities with bans and cities without them — the difference is civic engagement. GDPR and data rights: The European General Data Protection Regulation, which took effect in 2018, established broad rights for individuals regarding how their data is collected, stored, and used. It was the product of a decade of advocacy by privacy researchers, civil society organizations, and European parliamentarians — and it has influenced data practices globally, including in countries where the regulation does not legally apply, because multinational companies find it operationally easier to apply high standards uniformly. Public comment on AI regulation: In 2023, the U.S. Copyright Office, the FTC, and multiple other agencies solicited public comments on AI policy questions. Tens of thousands of individuals, organizations, and companies submitted comments. These comments are part of the official regulatory record and influence final rules. Individual citizens with domain expertise — teachers, doctors, artists, journalists, workers — have submitted comments that shaped regulatory outcomes.

These examples share a structure: a group of people who cared about a specific outcome, understood the governance mechanism relevant to that outcome, organized effectively to engage that mechanism, and changed what AI is permitted to do. None of these people needed to write a neural network. They needed domain knowledge, civic knowledge, and the ability to communicate effectively.

The Relationship Between Technical Literacy and Civic Power

One of the structural challenges of AI governance is asymmetric expertise: the people with the deepest knowledge of how AI systems work are often the people building them, who may have direct financial interests in particular policy outcomes. Citizens without technical backgrounds find it difficult to evaluate competing claims about AI capabilities, risks, and limitations — and can be manipulated by those with more information. This is why technical literacy matters for civic engagement. You do not need to know how to train a transformer architecture to participate effectively in AI governance. You do need to be able to: Evaluate accuracy claims: When a company says its AI system achieves 95% accuracy, you need to ask: on what population, measured how, compared to what baseline, and with what consequences for the 5% errors? Identify affected communities: You need to be able to ask who is affected by this system that is not represented in this room. Recognize opacity as a governance problem: When a company or government agency uses an AI system to make consequential decisions but refuses to explain how it works, you need to recognize that opacity as something citizens have legitimate interests in challenging — regardless of whether the system is technically sophisticated. These are analytical skills, not engineering skills. They are the skills this track has been building.