Case Study

Reimagining Policy Design through top-down Societal Optimization Approach


The Australian Federal Government is Australia’s highest level of government  responsible for the conduct of national affairs.

Its areas of responsibility are stated in the Australian Constitution and include defense and foreign affairs; trade, commerce and currency; immigration; postal services, telecommunications and broadcasting; air travel; most social services and pensions.

CSIRO is Australia’s peak federally funded body that is advancing scientific research and innovation in the national interest.

CSIRO is among the world's largest and most diverse scientific research organizations, dedicated to addressing Australia's most pressing challenges - from food security, climate change, defense and artificial intelligence.

The Challenge

As AI continues to rapidly evolve, it has become a leading issue for sovereign capability. The Australian Federal Government/CSIRO has identified AI as a critical technology.

Ensuring the safe deployment and usage of AI is high on the agenda for not just the Australian Federal Government/CSIRO, but sovereign nations around the world.

Solutions and policies currently focus on creating guardrails and mitigation of bias in training data.

While these methods are useful in addressing today’s issues such as censoring unwanted outputs from large language models, they are not useful in ensuring that AI will be fundamentally, and intrinsically safe.

The Australian Federal Government/CSIRO was in need of a comprehensive approach to the safe and responsible development of AI to guide the national interest.

How Akin Helped

Akin’s was part of the early working group brought together to informing the development of this policy framework disability.

Akin management was was also called upon to meet with the Australian Federal Task force on Generative AI and Responsible AI, and met with the Sustainability commission for the Federal Disability Health insurance agency.

Akins approach to the safe and responsible development of AI is ensuring intrinsic alignment - that is raising AI to have values and behaviours that are designed around the simple premise “Is the life of the person or ecosystem using the AI improving?” Thai ensures a core focus on outcomes optimization.

This framework was published in 2023 as the Safe and responsible AI in Australia discussion paper.

Australia has emerged with a unique and scalable approach the responsible use of AI, Rather than the primary focus being in data management and guardrails, the first guiding principal is

“Human, societal and environmental wellbeing: AI systems should benefit individuals, society and the environment."

This is a scalable and future-proof approach, which focuses attention on positive outcomes optimization

There is also practical implementation policy in place, for example the field use of AI in high-compliance sectors like health and disability support.

Although Akin did not directly develop this policy, we are proud of our work as early pioneers in Responsible AI (as one of the first AI Public Benefit corporation to be formed), in having a strong voice in policy formation, and in developing practical tools for real-world optimization and responsible co-evolution methodologies.

“What Akin’s technology is doing is inspiring, spotlighting the massive potential for AI to improve the quality of life for people in ways that weren’t possible before. It also shows how assistive technology, levering off AI, can play a crucial role in sectors challenged by skill shortage.”

Hon Ed Husic MP
Minister for Innovation, Australian Federal Government