The Department of Employment and Workplace Relations is trialling the open use of ChatGPT for its staff, marking a notable exception to the broader Australian government ban on artificial intelligence tools. The trial, which began earlier this month, allows employees to use the generative AI chatbot for a range of work-related tasks, including drafting documents, summarising information, and generating ideas.
Contrast with wider government policy
The move comes despite a directive from the Australian Signals Directorate and the Department of Home Affairs that prohibits the use of AI tools like ChatGPT on government devices due to cybersecurity risks. The ban, announced in March, applies to all federal government departments and agencies, citing concerns about data privacy and the potential for sensitive information to be shared with third parties.
However, the workplace department has received an exemption to conduct the trial, which is being closely monitored by the Digital Transformation Agency. A spokesperson for the department said the trial is part of a broader strategy to understand the potential benefits and risks of generative AI in the public service.
Scope of the trial
Under the trial, staff are permitted to use ChatGPT on work-issued devices but are instructed not to input any classified or sensitive information. The department has also implemented additional security measures, including data anonymisation and regular audits. The trial will run for three months, after which the department will evaluate the results and decide whether to expand the use of AI tools.
“We are taking a cautious but proactive approach to exploring how generative AI can improve productivity and service delivery,” the spokesperson said. “We are also keenly aware of the risks, and we are working closely with cybersecurity experts to mitigate them.”
Reactions from experts and unions
The trial has drawn mixed reactions from technology experts and public sector unions. Dr. Sarah Johnson, a cybersecurity researcher at the University of New South Wales, said the trial is a positive step but must be handled carefully. “It’s important for the government to understand these tools, but they cannot afford to compromise security,” she said. “The trial should be a model for how to integrate AI safely.”
Meanwhile, the Community and Public Sector Union has expressed concerns about potential job displacement and the need for proper training. “We want to ensure that any use of AI is done in a way that supports workers, not replaces them,” a union spokesperson said.
The trial is being watched closely by other government departments, some of which have expressed interest in similar initiatives. The Digital Transformation Agency is expected to release guidelines for the safe use of AI in government later this year.
