Concern 201 · Working draft · editorial review in progress
Training-data secrecy
Training-data secrecy is limited visibility into the material used to train a model and the choices made when collecting it.
Understanding the concern
Without a useful account, researchers and affected communities may struggle to examine representation, privacy, or rights concerns. Developers may also have legitimate reasons not to release personal data or confidential material. The challenge is providing meaningful transparency without creating further exposure.
Transparency does not always require publishing every training item. Summaries, independent review, and controlled access can answer different questions. Requirements vary: the EU has introduced a template for covered general-purpose model providers to summarize training content, with scope and transition rules.
An example
A publisher describes its research process and sources without releasing every private interview note.
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.
- EU
- The European Union, a group of European countries with shared institutions and some shared laws.
Evidence and its limits
The EU template describes a specific disclosure regime; the provenance audit explains documentation gaps. Neither implies a worldwide duty to publish full datasets.
European Commission · 2025 · Official regulatory guidance
Template for general-purpose AI model providers to summarise their training content ↗ (opens in a new tab)Describes public training summaries for covered providers, including information gaps and transition conditions.
Longpre and colleagues · Dataset audit
The Data Provenance Initiative: A Large Scale Audit of Dataset Licensing and Attribution in AI ↗ (opens in a new tab)Traces dataset lineage, licensing, and attribution problems in audited collections; not proof that every third-party dataset is unlawful.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
What information would allow meaningful scrutiny while protecting people whose data appears in the collection?
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.
Trustworthiness
Trustworthiness means giving people good reasons to rely on you. It grows when promises, actions, and evidence line up. With AI, a friendly voice or a famous brand is not enough: people need to know what was checked and what could still go wrong.
In practice, keep promises small enough to test, record mistakes, and tell people when the tool's limits affect them. If a service promises human review, make sure a person really has time and authority to review it.
Ask together: What would give the people affected a sound reason to trust this use?
Accountability
Accountability means being answerable for decisions and taking responsibility for repair. It requires more than saying that the software made the choice.
Name the person or team responsible, keep useful records, and give them authority to correct problems. Make the route to a complaint clear to the people affected.
Ask together: Who will answer for the outcome and help put things right?
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?
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.