Research map / Large language models / AI and society
Position: The Right to AI
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
A position paper arguing for a "Right to AI" that gives individuals and communities meaningful participation and power in AI development and governance, grounded in Lefebvre's Right to the City and Arnstein's ladder of participation.
Problem
AI development and governance are concentrated in expert, corporate, and governmental hands, marginalizing affected communities and raising bias, legitimacy, epistemic autonomy, and data-ownership concerns.
Approach
Reconceptualizes AI as societal infrastructure and asserts a collective "power right" to co-shape its objectives, data practices, and oversight, operationalized as a four-tier model (consumer-based, private organization-led, government-controlled, citizen-controlled) adapted from Arnstein's ladder.
Key results
- Reviews nine participatory AI case studies (e.g., WeBuildAI, PRISM Alignment Dataset, Māori Data Sovereignty Initiative) and finds that most are context-specific and many remain at tokenistic rungs of participation.
- Proposes a four-tier Right to AI model that maps current AI governance paradigms from consumer-based to citizen-controlled.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use a participatory algorithm designed collaboratively with stakeholders | WeBuildAI · Donation allocation | equity | Helped |
| Involve patients, clinicians, and ethicists in iterative design | Co-Design of Trustworthy AI in Healthcare · Healthcare | diagnostic bias | Helped |
| Use human-in-the-loop feedback | Project Dorian · Crisis logistics | resource allocation speed | Helped |
Surprising findings
- Reports that WeBuildAI's community-designed matching algorithm produced more equitable donation allocations than traditional human-led methods, while participation can also be tokenistic or co-opted.
Limitations
- Conceptual argument without empirical validation of the four-tier model.
- Case studies are drawn from prior literature and are heterogeneous, so lessons may not generalize.
- Acknowledges scalability, funding, and institutional barriers to citizen-controlled governance.
Experimental setup
Builds on: Right to the City Arnstein's ladder of citizen participation Design Justice Ostrom's commons governance
Closest papers by meaning
- Position: Democratic AI is Possible. The Democracy Levels Framework Shows How It Might Work. (ICML 2025 · similarity 0.82)
- Towards a Theory of AI Personhood (AAAI 2025 · similarity 0.82)
- Position: Generative AI Regulation Can Learn from Social Media Regulation (ICML 2025 · similarity 0.82)
- Position: Technical Research and Talent is Needed for Effective AI Governance (ICML 2024 · similarity 0.82)
- Position: If Innovation in AI systematically Violates Fundamental Rights, Is It Innovation at All? (NeurIPS 2025 · similarity 0.82)
- Resource Democratization: Is Compute the Binding Constraint on AI Research? (AAAI 2024 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.