Labor has said that if it assumes office at the next election that any Commonwealth automation initiatives will be undertaken under close human supervision.
The comment was made during press rounds on Wednesday following the delivery of the 2019-20 federal Budget. Shadow Minister for Families and Social Services Linda Burney, alongside Shadow Treasurer Chris Bowen, responded to questions on the government's funding boost for the Single Touch Payroll (STP) system.
The Australian Taxation Office (ATO) and the Department of Veterans' Affairs were given AU$82.4 million in funding for the next four years to expand data being collected through the STP program.
As of July 1, 2020, income support recipients will report fortnightly on their income, with the data to be shared with the Department of Human Services (DHS) -- the government agency that executed the data-matching initiative that colloquially became known as "robo-debt".
See also: Australian Budget 2019: Everything you need to know
While agreeing with it in principal, Bowen said the STP program should be "applied very carefully".
"This government has an appalling track record on robo-debt so we would ensure that we put a human review and activity to ensure that this was applied consistently," he said.
"Obviously if a payment is made incorrectly, it should be dealt with and STP -- on a whole range of matters, not just social security -- provide government with new options going forward to ensure compliance and better administrative arrangements."
The Labor pair instead proposed that it would execute the program with "appropriate" checks and balances.
"This government's track record on robo-debt doesn't fill us with a great deal of joy, so we would ensure that there's appropriate check and balances to ensure that those Australians in receipt of payments -- who may or may not received one in error -- are dealt with dignity and appropriately," Bowen continued.
"Let's hope that this is not robo-debt mark two," Burney added.
"This government ... ignores the fact that the system has become so de-humanised, so difficult that many people are not able to access it.
"We will make sure that if we are afforded government that any of these issues will have a human eye overseeing automation. That is our commitment and it is a very important one for people who are welfare recipients."
In 2016, DHS kicked off the data-matching program of work that saw the automatic issuing of debt notices to those in receipt of welfare payments through the country's Centrelink scheme.
The Online Compliance Intervention (OCI) program automatically compares the income declared to the ATO against income declared to Centrelink, resulting in debt notices -- along with a 10 percent recovery fee -- being issued when a disparity in government data is detected.
One large error in the system was that it incorrectly calculated a recipient's income, basing fortnightly pay on their annual salary rather than taking a cumulative 26-week snapshot of what the recipient had paid.
As of June 2018, a total of AU$375 million had been spent by DHS on the contentious data-matching project.
Human Services in February also disclosed wiping, or writing-off, 26,104 OCI debts since July 2016. Additionally, 14,621 debts were reduced to zero.
In total, the department had initiated 925,728 assessments, and raised 409,572 debts from July 1, 2016, through to October 31, 2018. The average debt, the department said, was AU$2,184.
The Australian government and the loose definition of IT projects 'working well'
Straight-faced, a Department of Human Services representative told a Senate committee its data-matching 'robodebt' project went well, because it produced savings.
Why Australia is quickly developing a technology-based human rights problem (TechRepublic)
Human rights advocates have called on the Australian government to protect the rights of all in an era of change, saying tech should serve humanity, not exclude the most vulnerable members of society.
SAS recommends government start small with AI and machine learning
The company's director of Global Government Practice has said it's best to start small and build confidence in-house before pushing a data analytics-based project out to the masses and hoping for the best.