EXCLUSIVE: All but two Australian public service agencies required to appoint an existing staff member as their chief AI officer (CAIO) have met a government deadline to do so by the start of the new financial year.

Agencies were directed to appoint a CAIO by July 2026 under the public service AI Plan released by the Albanese government in November 2025, to help drive adoption of the technology across government.

Out of 106 non-corporate commonwealth entities required to appoint a CAIO, only two did not do so by 1 July, the Department of Finance confirmed to Information Age.

One was the Inspector-General of Taxation (IGT), which is expected to advise the department of its CAIO appointment later this month.

The other was the National Environmental Protection Agency (NEPA), which was only established as an agency on Wednesday, 1 July.

However, the other new agency established this week – the Australian Energy Regulator – did manage to appoint a CAIO by the deadline.

The Department of Finance also confirmed no commonwealth companies – which include organisations such as NBN Co, Snowy Hydro, and the National Australia Day Council – had advised them of a CAIO nomination.

Commonwealth companies and corporate commonwealth entities are only “strongly encouraged” to appoint a CAIO, but are not mandated to appoint one or to inform the department.

Around 24 per cent of corporate commonwealth entities (18 out of 74) had confirmed a CAIO nomination of their own, Finance added.

While only 56 Australian Public Service agencies and corporate entities had appointed a CAIO by 15 June according to figures published by the department, that number more than doubled to 113 by 29 June as government entities rushed to name a responsible individual.

AI chiefs not expected to be technical experts

Government agencies were told their nominated CAIO did not need to be a technical expert, and Information Age understands staffers given CAIO responsibilities will not be awarded extra pay.

CAIOs were described as “people whose primary job is pushing opportunity, working alongside those with more cautionary approaches and concerns” in a November 2025 information pack developed for agencies by the Department of Finance.

“They will need to be(come) comfortable with engaging with, for example, CIOs [chief information officers] and CISOs [chief information security officers] and learning about generative AI, its applications, and the associated technical concerns that might arise,” agencies were told.

“CAIOs need to be able to act as a port of call when the innovative implementation of generative AI hits fear, uncertainty and doubt.

“It is not about ignoring concerns, rather balancing them with the potential upsides and recognising safe and responsible adoption involves learning, experimentation and strategic nous.”

Agency CAIO appointments come as the federal government makes AI a key responsibility for public service agencies, which all previously appointed a so-called AI Accountable Official focused on monitoring risk and governance elements of the technology.

Smaller agencies were allowed to have both CAIO and AI Accountable Official roles covered by the same staff member, but the government preferred them to be separate.

In June, an audit revealed government agency IP Australia appointed its deputy director general as both its CAIO and its AI Accountable Official because it did not want to create “a dichotomy between driving adoption and engagement with AI and being accountable for governance and risk of AI use by appointing two different officials”.

Agencies were told by the Department of Finance that staff acting as both a CAIO and an AI Accountable Official “need to manage any tensions arising from being both advocate and guardian”, and that “expectations and strategies” for this should be “made explicitly clear”.

While the government’s Digital Transformation Agency (DTA) was expected to begin holding meetings with agency CAIOs earlier this year, Information Age understands this has yet to begin – however the DTA has held meetings with AI Accountable Officials.

Addressing ‘patchy’ AI leadership

Finance Minister Katy Gallagher told a Senate Estimates hearing in late May that AI leadership across agencies had been “patchy”, and the appointment of CAIOs sought to “drive that leadership and attention to AI, which was apparent in some agencies – but not all”.

“As [AI] becomes an increasing role in the delivery of public services, which it will over time, I just think we need to build that capacity – and make it a departmental responsibility,” she said.

“... You have to treat this like your CIO or your COO [chief operating officer], and we need senior people to take it on.

“They need to be a point of contact, and they need to be driving it and responsible for it within their agency.”

Department of Finance secretary Matt Yannopoulos told the same hearing that appointing CAIOs was “not about distributing governance, or diluting it”.

"It’s about trying to find people who are going to provide an advocacy and engagement point in each of these entities, for my team, to promote safe and ethical application of this technology, and improve efficiency for services for Australians,” he said.

Public service CAIOs have access to senior leadership training programs through the Australian Public Service Commission (APSC), which manages and supports public sector workers.

The APSC is also considering further specific training options for CAIOs in consultation with the Department of Finance and the DTA.

Finance’s AI Delivery and Enablement (AIDE) function "will convene and work with the CAIOs to support them in driving adoption within their agencies and across the APS", according to the department’s information pack for agencies.

“CAIOs will be supported by early adopters and experts within and across agencies who have hands-on experience applying AI in their work, whether for personal efficiency, specific job-related functions, or workflow integration,” it stated.

More than 100,000 of Australia’s almost 200,000 public service staff have completed mandatory AI foundation training, Information Age understands.

Dozens of government agencies missed a February 2025 deadline to publish transparency statements about their use of AI, but DTA official Lucy Poole said all relevant agencies had now complied, “with a further 21 agencies voluntarily publishing statements”.