Pattern library

Surface the key factors that influenced an AI decision in a way that is readable, appropriately qualified, and supports the user's ability to evaluate, challenge, or act on the outcome.

Surface the AI system's confidence level and uncertainty range in a way that is immediately visible, readable by non-technical users, and calibrated to the actual uncertainty in the output: preventing both over-trust and unnecessary alarm.

Help users understand an AI decision by showing how it relates to comparable cases: building intuition through analogy rather than abstraction.

Communicate the intended operating scope of an AI system, and signal clearly when a specific case, query, or context falls outside it: so users can make informed judgments about whether to rely on the output.

Give users visibility into the origin, quality, and appropriateness of the data used in an AI decision: at both the model training level and the specific decision level.

Provide authorized users with clear, accessible mechanisms to review, modify, and override AI recommendations, and ensure that all interventions are documented in a way that supports accountability and audit.

Organize explanatory content into meaningful depth layers: so the right amount of information is available to each user at the right moment, without overwhelming those who need less or frustrating those who need more.

Make the presence and role of AI visible to users at the relevant point of interaction: enabling them to calibrate their trust, seek human alternatives where needed, and exercise their rights in relation to automated processing.

Provide a clear, chronological view of the key events in a decision's lifecycle, including data collection, model assessment, human review, overrides, and communications, in a form that supports both user understanding and audit.

Provide a clear, accessible, and functional mechanism through which the person affected by an AI-influenced decision can formally challenge it: with confidence that the challenge will be received, reviewed by a human, and result in a documented response.

Translate the factors behind an adverse AI decision into specific, realistic, controllable next steps: giving users a genuine pathway to a different outcome rather than a list of features to optimize.

Provide a structured, accessible, and honest documentation of an AI system, covering its purpose, data, performance, limitations, fairness properties, and governance, in a form that serves operators, affected persons, regulators, and the public.

Adapt the content, depth, language, and available actions of an explanation to match the specific role, authority, and information needs of the person viewing it.

Provide an interactive mechanism through which users can ask follow-up questions about a decision or output, probe specific factors, test hypothetical changes, and receive direct answers: without requiring technical knowledge.

Capture a comprehensive, tamper-evident, and accessible record of AI decisions, including inputs, outputs, model versions, human interventions, and data states, sufficient to support retrospective audit, regulatory review, and the exercise of user rights.

Provide a low-friction mechanism for users to report errors, provide corrections, and signal disagreement with AI outputs, and close the loop by communicating that feedback was received and, where appropriate, how it was used.

Show users the minimum or most relevant changes to their situation that would have produced a different outcome: making the decision boundary visible without requiring users to understand the model.

Surface indicators of how equitably an AI system performs across different populations: giving users, operators, and auditors the context to assess fairness concerns, and making bias a visible and contestable property rather than an invisible assumption.

Make personal data use visible, comprehensible, and genuinely consented to: giving users a clear understanding of what data is used in AI systems, for what purpose, and what they can do about it.

Surface the documents, passages, or records the system retrieved at runtime, so users can verify the basis of a response, assess its reliability, and access primary sources directly.

Give users visible signals about the factual grounding of AI-generated content: distinguishing between responses that are well-supported by retrievable evidence and those that may contain fabricated, outdated, or unverifiable claims.

Make the actions taken by an autonomous AI agent legible, reviewable, and auditable: before, during, and after execution: so users maintain meaningful oversight even when the agent operates without step-by-step human approval.

Signal to users when a model's reliability may have degraded since deployment, due to changes in the data environment, shifts in population characteristics, or elapsed time since last validation, so that trust in the system reflects its current performance rather than its historical evaluation.