Automated Decision-Making Reform

Closes 15 Jan 2025

This service needs cookies enabled.

Consultation questions: Transparency and Safeguards

Use of ADM must comply with administrative law and the rule of law. As the Robodebt Royal Commission outlined, decisions made using automation must be defensible, and the automated systems used to increase efficiencies must not create further barriers for marginalised people to access help. Legal academics have argued that safeguards should be implemented at each stage of the automation process, including in system design and implementation, during decision-making, and after automated decisions have been made.

This section seeks views on what safeguards (if any) should be required:

  • before an automation is implemented
  • at the system level
  • at the decision level
  • post-decision.
How should the need for transparency about the use of ADM be balanced with the need to protect sensitive information about business processes and systems?
What transparency rules would be appropriate to build into the framework?
What pre-implementation safeguards should apply where ADM is intended to be used?
What system-level safeguards should be required, to ensure that ADM operates appropriately?
What decision-level safeguards should there be for persons affected by decisions made using ADM (for example, review rights)?
Should individuals be notified of the use of ADM?
If yes, should notification be required at a specific point in the decision-making process, or should flexibility be provided to agencies about the appropriate time to make a notification?
Should there be any exemptions to ADM safeguards? If yes, what exemptions should be included and why?
Should the safeguards be different depending on the risks associated with the use of ADM for a particular decision or administrative action?