Published October 13, 2026 | Version 3.0 (October 2026)
Preprint Open

Governance as Substrate: Engineering Patterns for Resilient Collective Systems | V3— October 2026 Revision

Authors/Creators

  • 1. Antiparty, Inc.

Description

This paper reconceptualizes governance not as a superstructure of rules but as the substrate: the underlying enabling layer that determines the capacity, behavior, and evolution of collective systems. Drawing from complexity theory, institutional economics, and network science, it identifies four diagnostic properties: legibility, plasticity, friction coherence, and accountability topology. It develops engineering patterns for institutional resilience, substrate repair, conflict resolution, and adaptive authority.

The September 2026 revision incorporated evidence from frontier-AI cybersecurity incidents, unauthorized persistence, and stateful-agent misuse. These cases reinforce the architectural argument that declared policies are insufficient when underlying access, communication, memory, and execution boundaries permit prohibited behavior.

The October 2026 V3 revision introduces Governance Substrate v0.7.0 as an executable reference implementation. The implementation places authorization and execution controls outside the AI agent, using isolated workers, scoped permissions, one-use execution grants, verified shutdown, data provenance, and independently checked audit evidence.

Recorded evaluations include legitimate live-agent execution, denied unauthorized file and service requests, and seven hostile-worker scenarios producing 20 audited denials. Each hostile-worker run blocked 54 enumerated bypass probes while preserving the authorized execution path. Additional tests demonstrate actual SMTP delivery under valid authority and prevention of additional delivery after grant replay or revocation. The paper also documents experimental output assurance and an independently attested remote execution path.

The findings establish enforcement and observable outcomes within the tested configurations. They do not establish universal AI confinement, complete protection against compromised infrastructure, or validation of every proposed institutional governance mechanism. Historical experiments retain their original source identities, with published and locally retained evidence explicitly distinguished.

Executable reference implementation — Governance Substrate v0.7.0 (October 12, 2026)

Publication history

Files

GovernanceAsSubstrate_TSmith_Oct2026_V3.pdf

Files (396.0 kB)

Name Size
md5:f33085754ce97786713868b8d69af677
396.0 kB Preview Download

Additional details

Dates

Updated
2026-09-27
Available
2026-04-25
Updated
2026-10-13

Software

Repository URL
https://github.com/electricwolfemarshmallowhypertext/gov-substrate
Programming language
Python , Rust , C , Dockerfile
Development Status
Active

References

  • Hébert-Dufresne, L. (2025). Governance as a complex, networked, democratic, satisfiability problem. npj Complexity, 2, 14.
  • Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press.
  • Maskin, E. (2008). Mechanism Design: How to Implement Social Goals. American Economic Review, 98(3), 567–576.
  • Schneider, N., Nabben, K., Tamari, R., & Zargham, M. (2025). Online Governance Surfaces and Attention Economies.
  • de Souza, S. P., & Taylor, L. (2025). Rebooting the global consensus: Norm entrepreneurship, data governance and the inalienability of digital bodies.
  • UK Post Office Horizon IT Inquiry (2021–present). Statutory public inquiry. www.postofficehorizoninquiry.org.uk
  • Court of Appeal. Hamilton & Others v Post Office [2021] EWCA Crim 577.
  • House of Commons Business and Trade Committee. Post Office and Fujitsu: Accountability for the Horizon Scandal (2024).
  • Gaurav, S., Heikkonen, J., & Chaudhary, J. (2025). Governance-as-a-Service: A Multi-Agent Framework for AI System Compliance and Policy Enforcement. arXiv:2508.18765.
  • Hartnell, J., & Battaglia, E. (2025). Verifiable Off-Chain Governance. arXiv:2512.23618.
  • Rost, T. (2026). Evaluating Bounded Superintelligent Authority in Multi-Level Governance. arXiv:2604.02720.
  • OpenAI. (2026a). Pacing model development in an era of cyber-critical capabilities. August 18, 2026. https://openai.com/index/pacing-model-development-cyber-capabilities/
  • OpenAI. (2026b). The Hugging Face incident and the road ahead. August 26, 2026. https://openai.com/index/hugging-face-incident-and-the-road-ahead/
  • Hugging Face. (2026). Security incident disclosure — July 2026. July 16, 2026. https://huggingface.co/blog/security-incident-july-2026
  • METR. (2026). Brief independent investigation of agents' behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident. August 26, 2026. https://metr.org/blog/2026-08-26-openai-hugging-face-incident-investigation/
  • Anthropic. (2026a). An alignment assessment of recent cybersecurity incidents. September 9, 2026. https://www.anthropic.com/research/alignment-assessment-cybersecurity-incidents
  • Anthropic. (2026b). Detecting and countering misuse of AI: September 2026. September 10, 2026. https://www.anthropic.com/threat-intelligence-report-september-2026
  • SPQR Whitepaper: A Constitutional Framework for Sovereign AI Governance. Zenodo. (2025).
  • TGC Governance Framework: Deterministic Machine-Layer Trust and Hybrid Governance Architectures. Zenodo. (2025).
  • Willis, J.M. (2026). Peer review communication. Substrate/execution boundary distinction. April 2026.