BackTransparency and explainability by design

This page is still being polished. If you have thoughts, please share them via the feedback form.

Data on this page is preliminary and may change. Please do not share or cite these figures publicly.

Transparency and explainability by design

ITechLaw (2019)|LLM classified

Mitigation Taxonomy

1AI System

1.1Model

1.1.4Model Architecture

Design-time architectural choices affecting safety, interpretability, and modularity.

Also in Model

1.1.1 Training Data1.1.2 Learning Objectives1.1.3 Capability Modification

Additional Information

4.1 Organisations that develop AI systems should ensure that the system logic and architecture serves to facilitate transparency and explainability requirements. In so far as is reasonably practicable, and taking into account the state of the art at the time, such systems should aim to be designed from the most fundamental level upwards to promote transparency and explainability by design. Where there is a choice between system architectures which are less or more opaque, the more transparent option should be preferred.4.2 Users of AI systems and persons subject to their decisions must have an effective way to seek remedy in the event that organisations that develop, deploy or use AI systems are not transparent about their use.

LLM Classification Details

Reasoning

Design-time architectural choices prioritize transparent, explainable system structures over opaque alternatives.

Code: 1.1.4Version: v0.5Classified: Jan 22, 2026

Part of

Transparency and Explainability

Organisations that develop, deploy or use AI systems and any national laws that regulate such use shall ensure that, to the extent reasonable given the circumstances and state of the art of the technology, such use is transparent and that the decision outcomes of the AI system are explainable.

Other mitigations from ITechLaw (2019) (35)

Ethical Purpose and Societal Benefit

Organisations that develop, deploy or use AI systems and any national laws that regulate such use should require the purposes of such implementation to be identified and ensure that such purposes are consistent with the overall ethical purposes of beneficence and non-maleficence, as well as the other principles of the Policy Framework for Responsible AI.

3.2.2 Technical Standards

Lifecycle:Plan and DesignActor:Governance ActorAIRM:Govern

Ethical Purpose and Societal Benefit > Overarching principles

2.1.3 Policies & Procedures

Lifecycle:Other (general)Actor:DeveloperAIRM:Govern

Ethical Purpose and Societal Benefit > Work and automation

2.2.1 Risk Assessment

Lifecycle:Operate and MonitorActor:UserAIRM:Govern

Ethical Purpose and Societal Benefit > Environmental impact

2.2.1 Risk Assessment

Lifecycle:Other (outside lifecycle)Actor:DeployerAIRM:Measure

Ethical Purpose and Societal Benefit > Weaponised AI