Without regulatory reform, automated decision-making bots risk undermining the integrity of government agencies, says a think tank.

In a new report – Computer says no – the Centre for Public Integrity says, while ADM technologies offer opportunities to improve government efficiency and service delivery, “they also present significant risks to the integrity of public administration”.

The CPI points to the Robodebt scandal as a demonstration of “how unlawful practices embedded through automated systems can cause widespread harm”. More recent concerns about automated aged care assessments “demonstrate that the risks remain”.

Government agencies are legally bound by a framework that upholds principles of accountability, fairness and transparency in decision-making. However, says the think tank, “the use of ADM creates new difficulties for each of these assumptions”.