CoChat

A Layered Model for AI Governance

Open access
Gasser, Urs, Almeida, Virgilio A.F. · 2017 · IEEE Internet Computing

The DOI matches its Crossref record; 16 of the 18 references it cites were found in a registry or an index; none of the 14 sources its claims cite is checked for support yet.

Abstract

AI-based systems are “black boxes,” resulting in massive information asymmetries between the developers of such systems and consumers and policymakers. In order to bridge this information gap, this article proposes a conceptual framework for thinking about governance for AI.

Overview

CoChat Summary

TL;DR

The paper proposes a three-layer conceptual framework for AI governance—technical, ethical, and social/legal—to address information asymmetries and other structural challenges. It argues that AI governance should move beyond state-centric command-and-control regulation toward modular, layered, multi-stakeholder approaches inspired by Internet governance.

Key findings

  • Proposes a layered governance model with three interacting layers: social and legal; ethical; and technical foundations that support the ethical and social layers, sitting between society and AI applications.
  • Identifies three structural challenges that translate into design requirements: information asymmetries, finding normative consensus, and government mismatches.
  • Argues governance instruments can be developed at different times: near-term standards and principles for AI algorithms, and mid- to long-term legislation for mature AI applications.
  • Maps specific elements onto layers: technical-layer principles for accountable algorithms (responsibility, explainability, accuracy, auditability, fairness) and data governance anti-discrimination; ethical-layer human rights and IEEE principles; social/legal-layer institutions and responsibilities, such as Matthew Scherer’s proposed AI policymaking body and the GDPR’s limited “right to explanation.”

Methods

  • Conceptual and analytical article; no empirical data, sample, or quantitative analysis are reported.
  • Develops the framework by synthesizing AI governance challenges, existing governance models (active matrix theory, polycentric governance, hybrid regulation, mesh regulation), and layered modularity examples (OSI reference model and David Clark’s four-layer cyberspace model).
  • Main analysis is normative and design-oriented: it translates structural challenges into design requirements for a future AI governance model.

Limitations

  • The paper states no formal limitations.
  • Evident from the text: the framework is presented as a heuristic, and the paper says the exact contours of a future AI governance model are still in flux; it does not empirically test or validate the model.
  • Evident from the text: no implementation details, metrics, or case-study evaluation are provided for how the layers would operate in practice.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1109/mic.2017.4180835

Cites

16 of 16 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 Valid · 0 Not valid · 14 Not checked · 2 Own results and interpretations

14 references still to check

A free account verifies the references that are still open.

Verify all 14 references — free account
  • Probably not a substantial claim. 14

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Citations

Cites

Show all 16 references

Research works printed but not linked: 6

  1. gasser, 2017, AI and the Law: Setting the Stage, Medium
  2. 2017, The IEEE Global Initiative for Ethical Considerations in Artificial Intelligence and Autonomous Systems, Ethically Aligned Design A Vision for Prioritizing Human Wellbeing with Artificial Intelligence and Autonomous Systems (AI/AS)
  3. 2016, Big Data: A Report on Algorithmic Systems, Opportunity, and Civil Rights, Tech Report
  4. levin, 2017, New AI Can Guess Whether You're Gay or Straight from a Photograph, The Guardian
  5. mccarthy, 1955, A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence
  6. brownsword, 2008, Regulating Technologies Legal Futures Regulatory Frames and Technological Fixes Hart

Cited by

Items 101–200 of 428

Metadata

Authors
Gasser, U., & Almeida, V. A. F.
Affiliations
Harvard University (US)
Venue
IEEE Internet Computing
Published
2017-11
Volume
21
Issue
6
Pages
58-62
ISSN
1089-7801 (ISSN-L)
OpenAlex
W2768432182
Cited by
431
References
18
Citations by year
2026: 69 · 2025: 100 · 2024: 57 · 2023: 56 · 2022: 61 · 2021: 42 · 2020: 29 · 2019: 13 · 2018: 4
FWCI
28.87
Type
journal article
Language
en
Open access
Locations
  • IEEE Internet Computing (publishedVersion)
  • Digital Access to Scholarship at Harvard (DASH) (Harvard University) (submittedVersion, cc-by)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
  • Social Sciences › Law › Digitalization, Law, and Regulation
  • Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
Keywords
AI governance, layered model, black-box systems, information asymmetry, conceptual framework, accountability, transparency, policy makers

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-05, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-05 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.