Concern 66 · Working draft · editorial review in progress
RAG data leakage
RAG data leakage occurs when an AI search-and-answer system gives a user information from documents they are not allowed to access.
Understanding the concern
The model may be functioning as designed while the surrounding retrieval system applies permissions incorrectly. A document copied into a search index must retain the relevant access limits; otherwise a summary can bypass the original document's locked door.
This differs from model memorization because the information may be fetched at the moment of the question. Asking the chatbot to keep secrets is not a substitute for enforcing permissions before protected material reaches it.
An example
An employee asks a routine question and receives an answer based on a restricted personnel file in a shared search index.
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.
- RAG
- Retrieval-augmented generation: an AI looks up material before writing an answer. Finding a document does not guarantee that the document or the answer is correct.
Evidence and its limits
Microsoft's documentation describes document-level authorization for retrieval systems; the example is hypothetical.
Microsoft Learn · Official product documentation
Document-level access control in Azure AI Search ↗ (opens in a new tab)Explains authorization around retrieval documents; architectural guidance is applicable context, not a claim that all services use the same controls.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
Are document permissions checked for this user before the material is retrieved and summarized?
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.