Concern 62 · Working draft · editorial review in progress
Membership inference attacks
Membership inference tries to determine whether a particular person's record or example was included in a model's training data.
Understanding the concern
The attacker seeks a yes-or-no fact about participation, not necessarily the record's full contents. That fact can be sensitive when the dataset relates to a health condition, a private service, or another revealing activity.
Membership claims can be uncertain or wrong. A successful study under one access setting does not prove that any outside user can make the same inference. Evaluation should examine what the model reveals and how much an attacker already knows.
An example
Learning that a name appears in a specialist clinic's training dataset could reveal something even without opening the medical record.
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.
- Inference
- Drawing a conclusion from clues. In AI engineering it can also mean running a trained model to produce an output.
Evidence and its limits
The original study defines and tests membership inference under specified access and knowledge assumptions.
Shokri and colleagues · IEEE S&P 2017 · Research study
Membership Inference Attacks Against Machine Learning Models ↗ (opens in a new tab)Defines and evaluates inference about participation in training under stated attacker access assumptions.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
Could the system expose who participated, even when it does not reveal their full records?
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.