It’s budget time again. While the government deals with ubiquitous cost of living pressures we have had a quick look to see what the budget holds for digital rights in 2025. And with an election likely to be called any day now, we plan on bringing digital rights issues to the government’s attention.
Here is what we found:
1. The OAIC has been handed an extra $8.7 million to support enforcement activities over the next 3 years. This is really a swings and round-a-bouts gesture–the funding replaces what was ripped out when dozens of jobs were slashed in response to a 23 per cent budget cut by the government in November 2024. They still need more funding and resources for education and community outreach. The OAIC will still be under resourced with this replacement funding and will be even more under resourced with Tranche II of privacy reform. On that note, wherefore art thou Tranche II privacy reform? It has completely vanished from the Canberra zeitgeist.
EFA calls on the government to publicly commit to delivering Tranche II of privacy reform in 2025.
2. Nothing for AI Safety. Zip. Zero. Zilch. This is bad. There was no funding or even an announcement of standing up an AI Safety Institute (or legislation) yet EFA notes Australia has signed on to the Seoul Declaration on AI safety and remains the last signatory to stand up an AI Safety body/regulator.
EFA underscores the necessity of a legal framework that ensures transparency, accountability, and fairness in AI systems. Such a framework must include:
- An Australian AI institute as per Australia signing on to the Seoul AI Summit Declaration in 2024.
- Mandatory risk assessments for AI applications in high-stakes domains like healthcare, law enforcement, and finance.
- Prohibited use cases for exploiting vulnerabilities, inferring emotions, biometric categorization, subliminal manipulation etc.
- Clear accountability mechanisms for developers and deployers of AI technologies including fairness, transparency and explainability.
- Strong privacy protections to safeguard individuals from intrusive data practices.
The absence of these guardrails not only jeopardises individual rights but also undermines public trust in AI technologies. EFA calls on the Australian Government to resume its leadership in AI regulation and set an example for the world by prioritising the safety and rights of its citizens over the rubbery, short-term economic gains posited by BigTech.
3. $60 million for cyber and digital safety is a good start but probably not enough given the government’s plans for both, e.g. Cybersecurity Strategy, Social Media Minimum Age, Digital ID, MyGov uplift, managing misinformation and disinformation etc. More detail (and dollars) are needed.
4. While the government has made progress in executing against the recommendations made by the Robodebt Royal Commission, hard questions remain in respect of the allocation of sufficient funding required to uplift IT systems, remediate unfair, inaccurate and obscure decisions made by algorithmic decision systems (see also our comments on AI safety) and to prevent a similar algorithmic systems like Robodebt being foisted, on Australia’s aged care system). Watch “I Am Not a Number” learn more about algorithms used in NDIS support planning.
Electronic Frontiers believes Australia can be a leader in balancing AI safety and innovation. We believe that our government should set a global example with responsible policies that benefit all of us.
If you’d like to support our mission, you can donate $20 today (or whatever you can spare!), set up a monthly contribution for ongoing impact, or become a member and have a say in EFA’s future. Every bit of support makes a difference and we truly appreciate it.
This blog was written by EFA’s Chair, John Pane.
(Image credit: Unsplash)