Researchers propose society-centered framework for judging AI governance by recourse, not principles
A new arXiv paper introduces a framework assessing whether people can get notice of AI-influenced decisions, challenge them, and see corrections, tested against North Carolina and 49 state documents.
In plain words
Researchers propose judging AI policy by a practical question: can a person find out that AI shaped a decision affecting them, contest it, and get it fixed? They tested this lens on North Carolina and other US states, finding that governments often publish AI information without giving individuals enforceable rights to challenge outcomes.
What was announced
The authors, writing on arXiv, argue that governments are regulating AI before they can measure its public effects, while existing research has catalogued outputs such as principles, strategies, registers, and audits without a shared language for evaluating policy as it forms. Their Society-Centered AI Governance Framework assesses instruments in any jurisdiction by three tests: whether a person can receive notice, challenge a decision, and see it corrected. It also distinguishes an agency's duty from a right an individual can invoke, and traces which later instrument must enforce each commitment.
Applying the framework, the team interviewed nine members of North Carolina's AI Leadership Council before and after the state's 2026 AI roadmap, and read 49 AI governance documents from 19 states using provision-level coding that combined model assistance with human review. For the state's own AI systems, they report visibility-first governance: visibility duties are in effect, notice exists only as an agency duty with no individual right, and AI-specific challenge and correction remain advisory. The roadmap assigns owners and dates to these missing protections, so the gaps are documented but not yet closed. The authors also flag "legitimacy before protection" as a testable risk, and note that recent federal preemption proposals target AI development, leaving recourse for state systems likely a state decision.
What it does not show
- The source is an abstract only; the full paper's methods, data, and peer-review status are not visible, and the findings are not independently confirmed.
For security and IT teams
- Scope
- The framework targets recourse for AI-influenced decisions in public-sector governance, evaluated on North Carolina and 49 documents from 19 US states; it is a research proposal, not a deployed policy tool.
- Status
- The paper appears on arXiv as a new announcement dated November 2026; the abstract does not state whether it has been peer-reviewed.
Sources
Source-support review completed 14 Nov 2026. This story was drafted with AI assistance and checked against retained source passages. Automated checking is not independent verification.
Coverage: arXiv: abstract.
Updates and corrections
No corrections so far.
First published 14 Nov 2026
New to the subject? AI from the Ground Up starts from the beginning.