Linkedin
Lucid Virtual Solutions
Schedule A Call

Australia’s new AI policy framework is now live. What you need to know.

On 1 September 2024, the Australian Government launched its policy on ethical use of Artificial Intelligence (AI). The policy represents the government’s lead in adopting AI for the benefit of Australians while ensuring its safe, ethical, and responsible use aligns with community expectations. Standards are being set in a three-phase approach to implementation.

Currently, the adoption of AI technology varies across the Australian Public Service (APS). The policy aims to standardise the government’s approach by setting baseline requirements on governance, assurance, and transparency. This will reduce obstacles to government adoption by boosting agencies’ confidence in their AI strategies and promoting safe and beneficial uses for the public good.

 

One significant barrier to successful AI adoption is public scepticism about how the government uses AI, including concerns about data use, transparency, accountability, and the impact of AI-assisted decision-making on individuals. The policy addresses these issues by implementing required and optional measures, such as performance monitoring, greater transparency about AI use, and standardised governance.

Artificial intelligence policy framework

What are some of the Policy Principles?

  1. Safely engage with AI. Improve productivity, decision-making, policy outcomes, and service delivery for Australians.
  2. Explain, justify, and own decisions. APS officers must explain, justify, and take ownership of AI-assisted advice and decisions.
  3. Clear accountabilities. Establish clear accountabilities for AI adoption and understand its use.
  4. Build AI capability for the long term. Develop and maintain AI capabilities within the APS for sustained benefits.

What are the mandatory requirements and recommended actions?

For government agencies, they must appoint accountable officials within 90 days of this policy’s implementation. Responsibilities may be assigned to individuals or the chair of a body and can be divided among officials or existing roles like Chief Information Officer or Chief Data Officer, based on agency preferences.

Accountable officials are responsible for:

  • Implementing this policy within their agencies.
  • Informing the Digital Transformation Agency (DTA) about new high-risk AI use cases
  • Serving as contact points for whole-of-government AI coordination.
  • Participating in government-wide AI forums and processes.
  • Staying updated with evolving requirements.


Private organisations and government agencies are strongly recommended to provide:

  • AI fundamentals training for all staff within 6 months of policy implementation, aligned with policy guidance.
  • Additional role-specific training for staff involved in AI procurement, development, training, and deployment.


Private organisations and government agencies should:

  • Identify where and how AI is used within the agency and develop an internal register.
  • Integrate AI considerations into existing frameworks such as privacy, security, record-keeping, cyber, and data management (
  • By prioritising these ethical principles and mandatory training requirements, Australia aims to establish a responsible AI ecosystem that benefits society and mitigates risks. These measures ensure that AI technologies are developed and used ethically, transparently, and accountably.

Why embedding the ethical principles of AI should be your top priority?

The ethical principles of AI aim to ensure that these systems are human-centred, fair, reliable, safe, and transparent, while protecting the privacy of the people they serve.

These principles are your compass. Whilst there’s been a lot of noise and disruption in the AI landscape, the principles of ethical and responsible use of AI haven’t changed. These should be your guide to making AI decisions at all levels and will support behaviours aligned to any internal policies and other approaches based on the mandatory obligations.

Contact us

References for this article

Share this Blog Post

Facebook
Twitter
LinkedIn

More Articles and Posts

accessible_learning_experience

An accessible course can still create an inaccessible learning experience

You can build an elearning course with strong colour contrast, meaningful alt text, accurate captions, keyboard navigation and screen-reader support, and still create an inaccessible learning experience.

The problem may happen before the learner even reaches the course.

The invitation email may contain unclear links. The login screen may not work properly with a password manager. Multi-factor authentication may require a code to be remembered and manually retyped. The course may be buried inside a Learning Management System (LMS) dashboard that is difficult to navigate without a mouse. It may launch in an unannounced pop-up, link to a scanned PDF or lose the learner’s progress when a session times out.

The course may be accessible. The experience is not.

This article is not another introduction to alt text, captions or colour contrast. Those things remain essential, and we have explored many of them in our earlier article on accessibility considerations in elearning design.

Instead, this article looks at the wider experience: how a learner receives, finds, launches, completes and receives evidence of learning—and where accessibility barriers can appear between each of those steps.

Read More

Already have AML/CTF training? What to update after the 1 July 2026 changes

The 1 July 2026 changes are now live for newly regulated sectors, including legal, accounting, real estate, conveyancing, precious metals and stones, and certain virtual asset services. For existing reporting entities, a number of reforms commenced earlier, on 31 March 2026, with transitional rules applying in some areas.

At Lucid, we’ve seen many organisations prepare for these changes by updating their existing AML/CTF courses rather than rebuilding them from scratch. Some have added new scenarios. Some have updated assessment questions. Others have refreshed facilitator guides, job aids, LMS records or role-specific modules.

But not every organisation has completed that work yet.

If your AML/CTF course was built before the reforms, it may still be mostly right. But there may now be gaps. It may refer to old processes, old terminology, old customer due diligence steps, or examples that no longer reflect the current regulatory environment.

This article is not an introduction to AML/CTF…. It’s an informal guide for organisations that already have training in place and now need to review what should be amended, added or refreshed following the July 2026 changes.

Read More