An open gateway for licensed AI access to engineering standards.
Engineering standards govern how the physical world is designed, built, tested, and maintained. They are licensed content, and most standards-developing organizations have responded to the rise of agentic AI with prohibition rather than access. Sèvres is the counter-argument: an open protocol and a platform that let an SDO serve licensed AI agents the same way it already serves human subscribers.
The name is borrowed from the Paris suburb that has hosted the Bureau International des Poids et Mesures since 1875. The metaphor is a vocabulary, not a claim: Sèvres is a gateway to authoritative references, not the place those references live. The standards live with the bodies that author them.
Specification
Clause-Level Access and Standards Protocol
An open specification that extends the IAB Tech Lab's Content Monetization Protocols (CoMP) for engineering and technical standards. Defines how an AI agent identifies itself, declares its intended use, and retrieves clause-level content with attached citations.
Read the spec at clasp.sevres.org →Platform
A CLASP-conformant gateway for standards bodies
A turnkey gateway for SDOs that want to publish AI-friendly access to their standards under CLASP. Hosts the CoMP front-door, the license registry, and the MCP retrieval layer behind a single deployment. Available as a self-hosted platform or a managed service.
See the demo at norma.sevres.org →Engineering SDOs have taken a defensive posture toward agentic AI. ISA is the most explicit:
ISA prohibits the entry of ISA standards and related ISA intellectual property (“ISA IP”) into any form of Artificial Intelligence (AI) tools, such as ChatGPT. Additionally, creating derivatives of ISA IP using AI is also prohibited without express written permission from ISA's CEO. Contact permissions@isa.org with any questions or requests. In the case of such prohibited uses, ISA will suspend the violator's access to ISA IP, and further legal action will be considered.— ISA Terms and Conditions, “Artificial Intelligence Usage.” isa.org/terms-and-conditions. Retrieved 2026-05-16.
ISA is not isolated. Other engineering SDOs have published comparable restrictions:
Citations verified 2026-05-18.
| Organization | Excerpt | Source |
|---|---|---|
| ASHRAE | “ASHRAE prohibits the entry of content from any ASHRAE publication or related ASHRAE intellectual property (IP) into any AI tool, including but not limited to ChatGPT.” | Standards & Guidelines page |
| ASTM | “Using Artificial Intelligence (AI) on ASTM standards and related intellectual property is prohibited. Violations will result in suspension of access.” | Digital Library banner |
| IEC | “You shall not use publications and documents sold through IEC Webstore for the development of any algorithm, digital model, software program, or AI tool.” | Webstore Terms § 3 |
| IEEE Xplore API | “... use any Content for the purpose of training, enhancing, developing, or contributing to any AI system, Machine Learning system, or Large Language Model.” (prohibited use) | API Terms § 2 |
| ISPE | “[You may not] use or enable artificial intelligence technologies and tools to ingest, train, test, analyze, process, copy, distribute, make publicly accessible, and/or generate output ... based on ... the Content.” | Guidance Document Portal T&C § 2.c.v |
The publishing industry has moved past this posture. The IAB Tech Lab finalized CoMP v1.0 on April 28, 2026, defining a clean handshake by which AI Systems and Content Owners can agree on commercial terms before content is accessed. Engineering SDOs are roughly two to five years behind that curve.
Prohibition is self-defeating. Practicing engineers increasingly work alongside agentic tooling and need to consult standards inside that workflow. An engineer blocked from consulting a standard at the moment of decision will work around the prohibition, ignore it, or simply not consult the standard. None of those outcomes serve the SDO, the engineer, or the field.
CLASP and Norma demonstrate that an SDO can preserve its subscription revenue and its IP protection while making content accessible to licensed AI agents. The architecture is strictly additive: existing subscription platforms continue to serve human users; new endpoints serve licensed AI agents. The revenue model is unchanged; only the access mode is expanded.