About the framework
Ethotechnics: governance and care infrastructure for high-stakes systems
Ethotechnics is an open framework for the operational accountability of high-stakes AI and socio-technical systems. It treats evidence, authorization, challenge, reconsideration, correction, and escalation as explicit system states — properties a system either has or lacks — rather than policy documents around the system.
The framework is authored by Kanav Jain, and developed through proposed standards, glossary curation, field diagnostics, and implementation-oriented research. Its standards are proposed, not imposed: each carries an editorial status (Draft or Stable), and adoption by institutions is what makes one authoritative.
Institute and Studio
How the Institute and Studio fit together
Ethotechnics Institute is the public, open-access commons: standards, ontology, glossary, diagnostics, and field notes are free to use, cite, and extend.
Ethotechnics Studio provides clinical AI safety evaluations, VC diligence, and governance advisory through commissioned engagements. The Institute publishes the open framework; the Studio applies it to specific systems and organizations.
What existing standards don't cover
The gap Ethotechnics addresses
NIST AI RMF tells you to manage risk. EU AI Act tells you to document compliance. ISO/IEC 42001 tells you to build a management system. OECD AI Principles tell you to be responsible. None of them tell you whether your system can actually be stopped when it's harming someone.
Existing frameworks govern by representation: documents, roles, reviews, and attestations. Ethotechnics governs by control: stop rules, binding clocks, rollback authority, and contestability that changes state.
| Existing standard says | Ethotechnics requires |
|---|---|
| "Maintain human oversight" (EU AI Act, Art. 14) | Named human with stop authority, tested halt path, recovery clock |
| "Manage risks across the AI lifecycle" (NIST AI RMF) | Measurable time-in-harm bounds, exercised rollback paths |
| "Conduct conformity assessment" (ISO/IEC 42001) | Evidence the system can be stopped mid-incident, not just documented as compliant |
| "Implement responsible AI principles" (OECD) | Binding escalation: owner + timer + action, or the system degrades |
Ethotechnics does not compete with these standards on whether governance is "important." It addresses what they leave undertreated: see the full comparison →
The governance landscape
Where Ethotechnics fits
AI governance is a rapidly growing domain. Enterprises invest in compliance platforms, evaluation tools, and guardrail systems to meet emerging regulatory requirements.
Ethotechnics is not a product. It publishes benchmarks and test cases that governance platforms can be evaluated against — the role NIST frameworks play for compliance tools, or OWASP standards for security tools. Those bodies hold that role because institutions adopted them. Ethotechnics is proposed for it.
The Institute provides the open standards. The Studio provides evaluations and diligence →
Foundations
Start with these core resources
- Glossary for core terms and definitions.
- Mechanisms catalog for decision logs, escalation patterns, and reversibility tools.
- Diagnostics for interactive self-serve tools.
- Standards for formal, versioned control definitions.
- Eval suites for governance benchmark suites.
- Field notes for applied essays and observations.
- Quick start guides for role-specific entry paths.
Citing this work
How to cite Ethotechnics
Ethotechnics Institute materials are published under CC BY 4.0 . To cite the framework:
Jain, K. (2026). Ethotechnics: A governance and care-infrastructure framework for high-stakes systems. Ethotechnics Institute. https://ethotechnics.org
Individual glossary entries and patterns include their own citation metadata.