Concern 290 · Working draft · editorial review in progress
CI/CD security for models
Security gaps in model delivery pipelines can allow an unverified or altered model to reach production.
Understanding the concern
CI/CD refers to automated processes that build, test, and release software. For AI, the pipeline can also move model files, training data references, and evaluation results. If these pieces are substituted or poorly linked, a release may not be the system the team intended to approve.
Signing and provenance tools already exist; calling them universally immature is misleading. The question is whether they are correctly applied to the full AI artifact chain. A valid signature identifies a source and integrity relationship, but does not establish that the model's behavior is appropriate.
An example
A sealed package is authentic, but it contains a different version from the one the review team tested.
Illustrative unless identified as a documented case.
Words in plain English
- AI
- Software that learns patterns from examples or uses rules to perform tasks such as recognizing images or generating text. It can make mistakes and does not understand every situation as a person would.
- Model
- The learned part of an AI system. It uses patterns found during training to produce an answer or prediction.
- Training
- The process of using examples to shape how an AI model behaves. Learning patterns from material does not establish that the material is true or fair.
- Provenance
- The history of information: where it came from and how it has been handled or changed.
- CI/CD
- Software practices for checking and releasing changes frequently. Automated releases still need suitable safeguards.
Evidence and its limits
The sources describe lifecycle integrity and attestation roles; neither treats cryptographic validity as behavioral safety.
NIST · SP 800-218A · 2024 · Official guidance
Secure Software Development Practices for Generative AI and Dual-Use Foundation Models ↗ (opens in a new tab)Adds AI-specific lifecycle practices to secure software development; guidance does not certify a particular system.
Internet Engineering Task Force · 2023 · Technical architecture standard
RFC 9334: Remote ATtestation procedureS Architecture ↗ (opens in a new tab)Defines evidence, verifier and appraisal roles; an authentic attestation does not establish that an AI evaluation is adequate.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
- NIST AI RMF (opens in a new tab)
framework
A question to discuss
Does the release process verify that the deployed artifact is the exact one approved by evaluation?
Keep track of different views, what needs checking, and any next step in your own discussion notes.
Open discussion notes →Principles for the conversation
Before you finish · reflect together
Virtues and principles to consider
Virtues are qualities we practice in how we treat people and make choices. They are not scores for judging someone. Use these ideas to explore what a thoughtful response could look like.
Stewardship
Stewardship means caring for resources and systems that others depend on, including people who will use them later. It looks beyond the benefit of a quick launch.
Plan for maintenance, energy use, data protection, and eventual shutdown. Ask who will look after the tool when the original team moves on and how future users can avoid inheriting hidden costs.
Ask together: What are we responsible for caring for over time?
Prudence
Prudence means using careful judgment before and during action. It balances worthwhile opportunities with the need to avoid harm that could have been anticipated.
Start with a limited trial, set stopping points, and prefer choices that can be undone when uncertainty is high. Review what happens rather than assuming the original plan remains sound.
Ask together: What is a careful next step, and when should we stop or change course?
Cooperation
Cooperation means working with others toward a shared purpose while respecting their contribution. It is not simply asking people to accept a decision already made.
Share information, roles, and credit. Include the people who do the daily work when choosing or testing a tool. Make disagreements discussable so that cooperation does not become pressure to agree.
Ask together: What can we do together that none of us can do well alone?
Editorial ratings
These are editorial judgments from the register, not measured probabilities or predictions about your situation.
About the register and scoring →More ways to explore
Search for more writing
For more specific or recent material, open a search below or ask for a web summary. AI-generated summaries can contain mistakes; open the original articles to check them.