Concern 221 · Working draft · editorial review in progress
Red-team disclosure norms
Red-team disclosure norms concern how to share findings from adversarial AI testing without unnecessarily enabling harm or concealing important risks.
Understanding the concern
A red team deliberately looks for failures under challenging conditions. Publishing findings can help users, researchers, and developers respond. Releasing detailed methods before a repair is available can sometimes make misuse easier, while indefinite secrecy can prevent accountability.
Different audiences may need different levels of detail at different times. A useful process records the finding, alerts responsible parties, sets a review timetable, and explains what is withheld and why. Disclosure should not be reduced to either publishing everything immediately or saying nothing.
An example
An inspector warns that a lock fails, coordinates a repair, and shares enough evidence for scrutiny without handing out an immediate break-in recipe.
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.
- Adversarial
- In this context, designed to challenge, trick, or attack a system.
Evidence and its limits
The sources support independent evaluation and vulnerability-response processes; they do not prescribe one disclosure timetable for every AI finding.
Longpre and colleagues · 2024 · Research proposal
A Safe Harbor for AI Evaluation and Red Teaming ↗ (opens in a new tab)Proposes legal and technical protections for independent safety research; the proposal is not itself binding policy.
NIST · SP 800-218A · 2024 · Official guidance
Secure Software Development Practices for Generative AI and Dual-Use Foundation Models ↗ (opens in a new tab)Adds AI-specific lifecycle practices to secure software development; guidance does not certify a particular system.
Broader frameworks and background
These resources provide context. They do not establish every claim in this concern.
A question to discuss
Who needs to know this finding now, and what timetable prevents temporary confidentiality from becoming permanent secrecy?
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.