AI Review Committee provides expert advice on governance

In Short:
– The AI Review Committee offers expert advice on AI use and reviews agency responses to serious incidents.
– The Committee emphasises governance, transparency, and oversight in AI projects within the government.

The AI Review Committee has established its terms of reference to offer expert, non-binding advice on AI use cases and review agency responses to serious AI incidents.At its first meeting, the Committee advised the Attorney-General’s Department and Services Australia on their upcoming AI projects.

AI governance framework

Chief Executive Officer of the DTA Chris Fechner noted the terms reflect the government’s commitment to strong governance amid AI adoption.

The terms emphasise transparency, continuous improvement and oversight for AI within government.

The Committee will review projects pre-deployment to help agencies determine appropriateness, risk identification, and safeguard adequacy.

Alignment with Australia’s AI Ethics Principles and governmental AI policies is crucial during assessments.

Post-incident reviews will focus on enhancing future governance, as stated by Mr Fechner.

External expert input may be sought when necessary, ensuring conflicts of interest are managed and security maintained.

The Committee will annually review its effectiveness, adapting to evolving AI technologies and governance practices.

Public trust is essential for AI’s success in government, emphasizing the need for transparency and continual refinement of approaches.

Two use cases from the Attorney-General’s Department and Services Australia were referred for Committee consideration.

The former aims to enhance staff productivity through AI-driven document analysis and workflow optimisation, while the latter tests AI for locating internal documentation.

Members appreciate the proposed productivity improvements but stress the importance of human oversight and robust quality assurance.

The Committee encourages ongoing focus on privacy, workforce capability, and planning for future organisational scale.

Both cases underscore that successful AI adoption hinges on technology and supportive organisational settings.

Committee responsibilities

Established under the AI Plan for the Australian Public Service, the Committee provides scrutiny of AI risks and safeguards.

It reviews high-risk AI use cases before deployment and incidents to identify lessons that bolster governance.

The Committee encompasses expertise in digital transformation, AI safety, privacy, cyber security, and workforce policies.

Accountability for implementation and decision-making remains with the agencies involved.

The Committee will convene as needed for case discussions and incident reviews.

For further details about the Committee, visit the AI Review Committee page.