The preferred supplier will undertake a rapid review and provide use cases on AI agents in the environmental regulation process. The preferred supplier will focus on external proponent/approval holder interactions with the regulatory process, instead of internal assessment processes managed by departmental staff. This includes, but is not limited to, document lodgement, communication and cost recovery payments.

The preferred supplier will identify and provide use cases where AI agent investment/integration could improve regulatory efficiency. This includes outlining the ethical considerations, cost, benefits, opportunities, limitations, governance, performance metrics and risk in relation to environmental regulation. Where possible, use cases must have regard to existing ICT infrastructure and compatibility with Dynamics 365. Furthermore, use cases must consider how AI agents can improve the transparency of regulatory processes and outline any information and data quality standards that are required to make this investment viable.

The preferred supplier must identify and outline the department’s capability gaps and provide an assessment on current staffing and technology infrastructure capability, and the governance requirements for data and model training. This will include considering and reporting on the extent to which data and information provided in referral, assessment and post-approval phases can be classified such that it is at least in some way accurate. The supplier must have regard to the level of autonomy and fragility of the AI agents in the use cases, and report ways to safeguard the regulatory regime from these risks.