Australian workers are seeking a greater say in how artificial intelligence (AI) is introduced and used at work, as more than half of Australian Services Union (ASU) members now expect the technology to significantly change their roles within the next three years.
The proportion of members anticipating major changes to their jobs has risen from 39.9% in 2025 to 51.3% in 2026, according to the union’s latest study.
The findings have prompted the ASU to call for greater transparency, worker consultation and a worker-first approach as organisations move beyond AI implementation.
Consultation gap
The rise in expectations around AI-driven workplace change comes as training and consultation remain relatively low among workers.
While AI use at work has increased by more than five percentage points over the past year, only 32.5% of surveyed members reported receiving AI-related training. Just 33.2% said they had been consulted specifically about the introduction or use of AI.
The ASU said workers can turn to AI secretly when employers place greater emphasis on productivity without providing adequate transparency, communication or guidance.
Almost half of respondents also ranked fear of job replacement among their top three concerns about AI. Other concerns included automation, surveillance, discrimination, outsourcing and the potential erosion of worker rights.
Worker voice
The ASU represents workers across occupations that could be significantly affected by AI adoption, including administrative and clerical workers, call centre employees, legal practitioners and IT professionals.
In its submission to the federal Joint Select Committee on Artificial Intelligence, the union has outlined five priorities for managing the digital transition:
Equitable access to AI and digital tools
Universal and accessible training for future jobs
Worker voice and strong safeguards
Recognition of essential skills
Sharing the productivity gains created by AI
ASU national secretary Emeline Gaske said workers need support to use AI effectively while ensuring the technology does not deepen existing inequalities.
“We represent a diverse workforce that is skilled and recognises that AI and new technologies can be a powerful tool to help with some of the more repetitive tasks at work, but they need the support to do that.”
“Our workforce is highly feminised, and so any disruption to their work from AI is likely to have gendered impacts across society, too.”
AI and job security
The union's member survey also found that nearly three-quarters of respondents believe workers should receive at least six months' notice if they lose their jobs as a result of AI-driven changes.
Members also called for stronger support so workers are not reduced to technology support roles as AI becomes more embedded in workplaces.
Gaske said workers and their unions should have a formal role in decisions around how AI is deployed and how the resulting benefits are distributed.
“AI has the potential to disrupt members’ jobs and rights, but it could also make work fairer, safer and more productive.”
“Realising those benefits requires workers and their unions to have a foundational, non-negotiable role in decisions about how AI is introduced and used, and how its benefits are shared.
“Workers must be at the heart of decision making to ensure protections are in place.
“If unemployment increases with no way for workers to upskill to harness AI tools, it is foreseeable that generations of workers will spend a lifetime without the quiet enjoyment of a career pathway, or will be unemployed or underemployed.”
Broader AI safeguards
The Australian Council of Trade Unions (ACTU) has separately called for stronger worker rights and AI that is pro-worker, safe and sovereign, alongside regulation of AI developers and deployers.
The ACTU has called for protection of Australians “from the unrestrained development and deployment of these systems in the Australian market” and proposed that the Artificial Intelligence Safety Institute oversee a licensing regime for the development and deployment of large frontier models and large open-weight models in Australia.
It has also called for legislated AI standards for data centre developments to mitigate their impact on communities.
The ACTU and AI academics have further called for investment in domestic AI capability and infrastructure, shared data centre computing capacity and a levy to fund Australia's local AI sector.
Submissions to the parliamentary committee closed on 14 September, with its final report due by 30 November 2026.