Concern 55 · Working draft · editorial review in progress
Chain-of-custody for training data
A training-data chain of custody records how material moves, changes, and is approved from its source to its use in a model.
Understanding the concern
This includes more than a source name. Teams may need to know who collected a file, which version was reviewed, what transformations occurred, and which dataset release used it. Such records help investigate a problem later.
The original definition implied that a cryptographic trail is always required. Cryptographic checks can support integrity, but they do not prove that material was truthful or lawfully obtained. The right controls depend on the task and risks.
An example
A signed delivery log can show that a package was unchanged in transit without proving that its contents are correct.
Illustrative unless identified as a documented case.
Words in plain English
- 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.
Evidence and its limits
W3C explains provenance relationships; web-poisoning research illustrates why versions and collection history matter.
World Wide Web Consortium · Technical standard
PROV Overview ↗ (opens in a new tab)Describes provenance records and relationships. Recording lineage does not guarantee the truth or lawful use of the material.
Carlini and colleagues · Security research
Poisoning Web-Scale Training Datasets is Practical ↗ (opens in a new tab)Studies mutable web sources and collection-time attacks. Demonstrated opportunities are not evidence that every dataset was actually attacked.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
Can we reconstruct what happened to this material, and what does that record actually prove?
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.
Truthfulness
Truthfulness means making claims that match the evidence and being clear about what we do not know. It includes correcting mistakes rather than protecting a polished story.
Check sources, label guesses, and distinguish a trial result from a proven benefit. If AI helped produce a claim, someone still needs to take responsibility for checking it.
Ask together: What can we honestly say, and what still needs checking?
Justice
Justice means fair treatment, fair procedures, and attention to who receives benefits and burdens. It asks whether people with less power can be heard and seek repair.
Look at outcomes across different groups and individual cases. Provide a usable way to challenge decisions and correct harm. Equal wording in a policy does not guarantee equal treatment in practice.
Ask together: Who benefits, who bears the cost, and who can challenge the decision?
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?
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.