Digital Rights Watch welcomes the opportunity to provide feedback on the Online Safety Amendment (Digital Duty of Care) Bill 2026.

The bill is an extremely welcome development, and DRW is broadly supportive of this initiative. As per our submission to the 2024 statutory review of the Online Safety Act:

“In principle, we agree with the intent of introducing a statutory duty of care for platforms; to flip the burden of responsibility of maintaining safety and security from individuals to corporations who require a social licence to operate in Australia. While we do not see it as a total solution to the problems that the online safety regime is seeking to address, if implemented well it can contribute to improving the situation.

There is an enormous imbalance of power between large technology companies, and the people who use their products and services.”

The Internet is a critical component of modern civil life and it is vital for the health of our democracy and of us as individuals that everyone is able to participate in the online world, free from harms from not only large tech corporations but also regular Australian businesses. We urge the Australian parliament to pass this bill expeditiously, as it offers the best chance to redress the power imbalance between technology corporations and regular people, and provide meaningful regulation of large tech platforms.

We think it is critically important that a human rights approach to online safety is central and would welcome more explicit commitments in this respect, including in the extraneous materials.

We do have some concerns about gaps in implementation and potential side-effects of some of the proposed legislation, which are addressed in our submission.

The bill could be greatly improved by considering the following:

Enforcement options must be improved, specifically to allow the duty to be directly enforceable by people who have been harmed.
Different levels of protection for children and adults will require privacy-invasive age assurance systems. It is not clear that this is justified.
Definitions of harm should be updated to better reflect the risks associated with generative AI chatbots and companions, as well as the needs of vulnerable people. Hate speech should be considered expansively, rather than attempting to define specific categories of hate speech.
The business models of tech companies are not contemplated in the framing of the bill, even though they are a significant driver of harm. Social media and other attention-economy business models give rise to the proliferation of extremist content, misogyny, scams, and deepfake material that is both harmful for individuals and our democracy.
The transparency provisions are a good start, but we need greater levels of transparency required of tech companies, including mandating detailed ad libraries.
Removal notices are a start but not enough: there ought to be strict liability for the publication and distribution of nudification content.