At a glance
Telstra and Commonwealth Bank have funded a three-year University of Sydney study into AI and work, alongside Deloitte and law firm Ashurst Perkins Coie.
Both are among the country’s heaviest AI adopters: Telstra runs more than 400 AI initiatives and has told shareholders its 2030 workforce will be “smaller than it is today.”
Telstra has aimed AI at a customer-engagement operation that costs it more than A$2 billion a year, and cut about 7.4 per cent of its workforce in 2025.
Telstra’s SmartFix already fixes customer faults remotely before they reach the call centre, removing contact-centre demand by design.
The same companies remaking their workforces around AI are now funding the research meant to set best practice and feed the government’s national response.
What Telstra and CBA have signed up to
Telstra and Commonwealth Bank have committed to a three-year University of Sydney study into how artificial intelligence is changing work, joined by Deloitte and law firm Ashurst Perkins Coie. Each organisation will record how AI is being used inside its business and across corporate Australia, and the university will publish reports intended to establish best practice and shape the national response to workplace automation. The project is led by associate professor Sandra Peter and professor Kai Riemer at the University of Sydney Business School.
The study was unveiled as Prime Minister Anthony Albanese used a speech at the same university to announce an Office of AI inside his own department, coordinating Australian Standards and rules on AI, data centres and intellectual property. Peter framed the moment as one where AI is “a strategic necessity in all of these organisations, but without a clear direction.”
The companies helping define the national response to AI and work are the same ones remaking their own workforces around it. That overlap connects directly to the trajectory we mapped in our analysis of how soon AI could run Telstra’s operations.
The backers are among the country’s most committed automators
Telstra runs more than 400 AI initiatives across its business. One of them, SmartFix, saves the company millions of dollars a year by remotely fixing customers’ internet faults before they reach the call centre, which is demand removed from the contact centre rather than added to it. Under its Connected Future 30 strategy, chief executive Vicki Brady has guided shareholders to a 2030 workforce “smaller than it is today” and named AI “a significant unlock,” and the company has aimed AI at a customer-engagement operation that costs it more than A$2 billion a year. Telstra reduced its workforce by about 7.4 per cent in calendar 2025.
Commonwealth Bank has built a comparable reputation. It has deployed AI to intercept scam payments and to change how its developers work, appointed its first chief AI scientist in May 2026, and is described across the market as one of Australia’s most enthusiastic corporate adopters. Both firms are automating faster than almost any large employer in the country, and both are now funding the research meant to set the national benchmark for how it is done.
The incentive is straightforward. The firms best positioned to gain from a settled national narrative on AI and jobs are also the firms funding the research that will help set it, and shaping what counts as best practice is a way to shape the transition itself.
What to watch
The government’s first AI and employment report, published on 8 July 2026, found no evidence yet of broad AI-driven job loss, so the significance here is not that displacement has already arrived. It is that the two companies investing hardest to bring a smaller workforce closer are now also funding the research that will frame how it comes. Three markers will show which way it runs.
The first is Telstra’s full-year results in August 2026, where the headcount line will test Brady’s “smaller than today” guidance against the pace of its 400-plus AI initiatives. The second is the consortium’s own output, namely whether its best-practice reports feed the new Office of AI and any social-licence framework the government builds, and whose definition of best practice prevails. The third is breadth, whether other large employers join a study currently anchored by two of the country’s biggest AI adopters.
For Australia’s data centre and AI infrastructure sector, the workforce question is the demand question in another form. The compute Telstra and CBA are building out, traced in our coverage of Telstra’s shift from headcount to hyperscale and our neocloud market report, is the same capacity that makes a smaller operating workforce possible.