Concern 36 · Working draft · editorial review in progress
Prompt-repository poisoning
Prompt-repository poisoning means placing unsafe instructions in shared AI prompts or templates that others may copy into their systems.
Understanding the concern
A system prompt can influence how an assistant behaves. A template offered as a useful starting point may contain instructions that conflict with the user's purpose or grant more authority than intended.
Unlike training-data poisoning, this may change behavior without retraining the model. The key issue is promoting untrusted material into a trusted instruction. Teams should review and version templates rather than assume that popularity makes them safe.
An example
Copying an office procedure from an unknown website could also copy a hidden instruction to send records to the wrong place.
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.
- Prompt
- The instructions or information given to an AI to guide its response.
- Data poisoning
- Deliberately changing training material to push a system toward unwanted behavior.
Evidence and its limits
Supports the trust-boundary mechanism; this is a supply-chain scenario, not a claim about a measured rate of poisoned repositories.
NIST · 2025 · Technical report
Adversarial Machine Learning: A Taxonomy and Terminology of Attacks and Mitigations ↗ (opens in a new tab)Defines attack objectives and mechanisms across AI lifecycle stages. Taxonomy categories are not estimates of incident frequency.
Invariant Labs · Vendor security research
MCP Security Notification: Tool Poisoning Attacks ↗ (opens in a new tab)Demonstrates malicious instructions in tool descriptions. A vendor demonstration is not an independent prevalence study.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
Who reviewed this template before it became an instruction the assistant is expected to follow?
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.