Artificial intelligence (AI) holds immense potential but realising it will depend on building public trust through safeguards that protect people.
AI is already changing how Australians work, learn, create and make decisions.
It can improve healthcare, lift productivity and help solve complex problems. But those benefits will only be realised if people are protected and can trust how the technology is developed and used.
Minderoo Foundation has worked on people-centred technology and AI policy for almost a decade, as the technology has become increasingly embedded in everyday life. During that time, we have supported research, brought experts and decision-makers together, and advocated for safeguards that protect people while allowing society to benefit from AI.
Here are eight things to know about how AI is evolving, what Australians are telling us and what comes next.
1
Why AI matters now
AI is becoming part of everyday life faster than the rules needed to govern it.
Used responsibly, it could make services better, businesses more productive and information easier to access. Without clear safeguards, it could also amplify scams and misinformation, weaken privacy and leave people with little recourse when automated systems cause harm.
Australia should not have to choose between innovation and safety.
As AI continues to evolve, people should be able to benefit from it with safeguards that reduce risk, protect their rights and ensure accountability when things go wrong.
2
Bringing evidence, experts and decision-makers together
The decisions Australia makes on AI policy need to be informed by evidence, independent expertise and the experience of people affected by the technology.
In late 2025, we convened SafeAI around a practical question: how can Australia capture the benefits of AI while putting the right safeguards in place?
SafeAI brings together evidence, international AI experts, civil society and decision-makers to advance a balanced approach to AI policy that protects people and enables innovation.
3
What Australians really think about AI
Through SafeAI, Minderoo commissions national track research to understand how Australians feel about AI, where they see value and what they expect from government and industry.
The latest survey of more than 2,000 Australians found that people can see the potential benefits, but many are cautious about the pace and the risks.
Almost two-thirds said AI is moving too fast, while 46 per cent believed the risks outweigh the benefits. Jobs, privacy, scams and misinformation remain among their biggest concerns.
Australians are not asking government to stop the technology. Most prefer clear rules that protect people while allowing innovation. But interestingly, when this balanced approach was removed and people were asked to choose between only minimal and strict rules, 69 per cent backed the stricter approach – even if it slows innovation.
Australians can see the opportunity, but their support will depend on governments and industry actively showing that the risks are being managed.
Research insights: Australian attitudes towards AI regulation
Minderoo Foundation commissioned a research study using different methods to get a clear picture of what people think about AI and how it should be regulated.
4
Bringing AI experts to Canberra
The research showed what Australians expect. The next step was putting that evidence in front of the leaders shaping Australia’s response.
In May 2026, SafeAI convened global experts in AI safety, governance and national security at Parliament House, Canberra.
They were joined by Deputy Prime Minister Richard Marles, Minister for Industry and Innovation Tim Ayres and Assistant Minister for Science, Technology and the Digital Economy Andrew Charlton, along with Members of Parliament and government officials.
Three priorities came through clearly. Australia needs:
- Visible safeguards that build public trust.
- Sovereign capability where national control matters most.
- A coordinated approach to AI policy across government.
Many of these themes were later reflected in the Prime Minister’s speech outlining Australia’s approach to AI.
The direction outline by the Prime Minister is welcomed. The next test is turning those principles into practical policy and ensuring safeguards move at the same speed as innovation.
Building trust in AI. Unlocking adoption.
AI Roundtable Outcomes Report.
Minderoo Foundation convened an expert AI roundtable at Parliament House in May, bringing together government representatives and domestic and international experts to discuss how Australia can scale AI safely and responsibly.
5
A fair deal for Australian creators
Roundtable participants also raised the need for regulatory clarity on copyright and the use of Australian content to train AI models. Echoing the Prime Minister’s speech, the consistent view was that any approach must give Australian creators a fair opportunity to benefit, and give responsible companies clear, workable rules to innovate.
Many of these themes were later reflected in the Prime Minister’s speech outlining Australia’s approach to AI.
The direction outline by the Prime Minister is welcomed. The next test is turning those principles into practical policy and ensuring safeguards move at the same speed as innovation.
Consolidated findings:
Public attitudes to AI
Data privacy, data centres and copyright. Prepared by Talbot Mills Research | July 2026.
Minderoo Foundation has a history of supporting creators through the challenges of a rapidly changing digital age. In 2022, we supported the establishment of the Public Interest Publishers Alliance (PIPA), a coalition of smaller and independent news organisations formed to collectively negotiate fair payment for the use of their content by Google under the News Media Bargaining Code.
That experience highlighted that smaller and independent organisations often lack the resources to negotiate alone and a collective approach can work for creators and responsible companies alike.
A fair and balanced solution matters for Australian artists. And attracting and building domestic AI capability gives Australia more levers to set safeguards to protect people and ensure the benefits are shared.
6
Making AI make sense
People should not need to be AI experts to understand how the technology can impact them.
Working with The Squiz podcast, we sought to bring SafeAI’s research and experts to a broader audience and explain the opportunities, risks and policy choices in an easy to understand digestible format.
The first podcast series explored AI harms, human rights and gaps in Australia’s regulatory system. After the Canberra roundtable, three further episodes looked at whether Australians trust AI, how AI agents could change the law, and what advanced AI means for national security.
The series gave people a clearer understanding of what is at stake and what responsible AI policy could look like.
Listen to the SafeAI episodes from The Squiz.
7
AI doesn’t stop at borders
AI is being developed across borders, so progress on safety must include the countries, companies and experts shaping its future.
During the World Artificial Intelligence Conference in Shanghai, Dr Andrew Forrest joined leaders to discuss how advanced AI can be developed safely and responsibly. Minderoo Founation and Tsinghua University also brought together senior officials, scientists and industry leaders from the United States, China and beyond to explore cooperation on shared frontier AI risks.
Minderoo joined the first UN Global Dialogue on AI and the AI for Good conference, in Geneva. Discussions focused on independent scientific evidence, child safety, human accountability and ensuring safeguards keep pace with the technology.
Some AI risks cannot be managed by one country acting alone. Minderoo’s role is to bring evidence into these discussions so countries, experts and organisations can work through shared risks and identify where cooperation is possible.
Our support for a new partnership between the Minderoo Centre for Technology & Democracy and the Cambridge Institute for Sustainability Leadership reflects this commitment.
By bringing together expertise in AI, technology governance and sustainability, the partnership will help strengthen governance, inform better policy and equip leaders to respond to shared challenges.
8
What’s next? The UN General Assembly
The next step is to take this evidence and international expertise into the United Nations General Assembly in New York.
Minderoo will focus on global cooperation on AI governance and safety. The opportunity is to connect governments, experts, civil society and philanthropy around practical safeguards and shared approaches that can continue beyond New York.
The focus will be on practical next steps, building common ground, connecting decision-makers with the right expertise and creating a clear pathway for further action.
Subscribe to our newsletter
Keep up to date on our Impact Missions, including AI, alongside our Communities, Natural Ecosystems, and Gender Equality work.
By signing up with your email, you agree to Minderoo Foundation’s Privacy Policy.