Australia National Framework for the Assurance of AI in Government (2024)
Cornerstones of assurance – Australia National Framework for the Assurance of AI in Government (2024)

Australia National Framework for the Assurance of AI in Government (2024) C.5: C.5 Procurement cornerstone: AI-aware procurement and contracts

Apply careful consideration to procurement documents and contracts for AI systems or products, which may cover AI ethics principles, clearly established accountabilities, transparency of data, access to relevant information assets and proof of performance testing across the system lifecycle. Keep contracts adaptable to technological change; build internal skills and knowledge transfer from vendors to avoid lock-in; carry out due diligence on new and amplified risks (for example the opacity of foundation models, privacy and security) and check whether standard contract clauses cover them; confirm the vendor can support review, ongoing monitoring and evaluation of outputs after an incident or a stakeholder concern, including evidence for review mechanisms; and weigh a component's benefits against its assurance challenges, prioritising alignment with the ethics principles alongside the desired outcomes.

Maintained by Gerard Blokdyk

What else in your programme already covers this

This control maps to 3 controls across 3 other frameworks. If you already hold one of them, the evidence you collected for it is the starting point here rather than new work.

  • 4.3.4 4.3.4 Put information requirements in supplier contracts

ISO/IEC 42001:2023 · 1 control

  • 5.4.2.f 5.4.2 f Management of third-party supplier risks

Every mapping shown was judged rather than inferred from wording similarity, and the ones that failed review are published too. See the coverage reports and what was rejected.

Other controls in Cornerstones of assurance – Australia National Framework for the Assurance of AI in Government (2024)

Query this from an agent

The graph holds this control, the 3 it maps to, and the evidence behind each claim, over MCP and REST.