Government warned over lack of transparency in algorithms as risk of another robodebt scandal looms
The federal government is under scrutiny for potential lack of transparency in using algorithms and artificial intelligence for decisions, which could lead to another robodebt situation, warns Dr Catherine Williams, CEO of the Centre for Public Integrity.
The Centre’s report, “Computer Says No,” calls for mandatory audits and a public register for all automated decision-making (ADM) and suggests independent monitoring over government use.
Concern arises from the government’s integrated assessment tool (IAT) for older Australians, found to under-assess needs, with aged care assessors unable to amend incorrect algorithmic decisions.
The ADM framework being developed by the attorney general’s department aims for safe ADM use, following recommendations from the robodebt royal commission.
ADM reform
Attorney General Michelle Rowland acknowledges ADM’s necessity for efficient service but hasn’t clarified the framework’s exact timeline or whether it will apply retrospectively.
In October, a report noted ignored transparency recommendations from the robodebt royal commission in the IAT’s development.
Independent MP Kate Chaney introduced a bill for a legislated ADM framework, criticising the government’s inconsistency and transparency issues.
Chaney emphasises the need for transparency in automated decision-making to ensure trust and clarity in government processes.
There’s pressure for the government not to abandon ADM framework development despite complexity, due to the need for accountability in decisions.
