The Australian Computer Society (ACS) has warned that the Federal Government’s proposed AI safety protections will fall short without stronger local technical capability to implement, test and enforce the measures.
In a statement responding to the Government’s AI safety package announced yesterday, ACS said legislation and safety frameworks “only work when capable, trusted technology professionals put them into practice”, arguing that baseline AI literacy across the public sector would not be enough to protect public infrastructure.
ACS said that when an AI agent gains unauthorised access to a government system, responding effectively requires specialist, locally resident expertise to determine what data or systems were accessed, assess broader risks to public services, and act quickly. It said similar expertise is needed for independent testing of powerful AI models and to evaluate safety claims made by developers.
The comments reference the “Rapid Review into the OpenAI incident involving Medicare”, with ACS arguing the Government’s six proposed safety protections and the review’s 18 recommendations do not sufficiently address the need for skilled personnel authorised to investigate and escalate AI risks inside the public service.
ACS CEO Dr Prins Ralston said: “Where are those people in the plan? How will we know they have the capability to do the job?”
ACS also cited research it said showed low public trust in AI developers, stating: “Only 23% of Australians trusted AI developers to ensure their software was safe (MIT and University of Queensland, SARA 2025, cited in AI & Data Centres submission, 2026).”
As part of its response, ACS called on the Government to set clearer capability and accountability requirements for roles involved in deploying and overseeing “high-impact” AI in government. It recommended assessing consequential AI roles against recognised skills frameworks such as SFIA, and requiring “named accountability” for State-used or State-funded AI, including a senior owner, a capable resident technology professional, and an officer authorised to pause or withdraw high-impact systems.
ACS also proposed that offshore developers training “covered large-scale models” should be required to appoint a qualified compliance officer based in Australia, and that the Australian Public Service workforce should be upskilled with specialist AI and cyber capabilities that go beyond basic digital literacy.
