Sustainability & AI
Algorithmic auditing
Independent examination of an algorithm's behaviour and impacts, extending assurance practice to AI systems.
Definition
The systematic, often independent examination of an algorithmic system's behaviour, impacts and compliance — testing outputs, documentation and process against defined standards.
Quick reference
At a glance
- Subject
- Sustainability & AI
- Editorial status
- Editorial draft
- Definition status
- Established
- Last updated
- 21 August 2026
References
This source provides part of the technical or institutional basis for the definition.
This source supports the explanation of how the term is applied, measured or governed in practice.
This source supports the wider sustainability significance and context described in the entry.
Overview
What it means
The practice extends social-science audit methods to software: probing systems with test cases, analysing outcome disparities and reviewing governance artefacts. A systematic review of the field (Bandy 2021) documents its rapid growth and persistent constraints — limited access to proprietary systems and inconsistent methods chief among them.
How it is used
Audits range from internal assurance reviews to external 'black-box' testing by researchers and regulators. The EU AI Act's conformity-assessment machinery and the emergence of AI assurance providers are formalising the practice; sustainability claims about AI systems are an emerging audit object.
Why it matters
As assurance professionals extend into AI, algorithmic auditing becomes the verification layer for AI claims — including environmental-performance claims — that markets and regulators can rely on.
Review
Help keep this definition useful and accurate.
Submitted reviews stay private until an editor decides whether to accept and attribute them.
Endorse this definition
Confirm what works
Endorse the current wording when it is accurate and useful. Suggested changes use the separate editorial form below.