Skip to main content

Australian AI and online safety proposals draw concerns over broad duties

15 NOVEMBER 2026·2 MIN READ·6 SOURCES

Ticker News reported concerns that Australia’s proposed online safety and AI rules could create uncertainty for technology companies and users. Bushnell argued that broad definitions of risk and harm could give administrators influence over how the rules are applied.

Key takeaways · 4

  • 01

    Assess how a digital duty of care could affect your service: the proposed obligation covers a broad range of potential risks, not only harms defined in legislation.

  • 02

    Factor in implementation uncertainty: Ticker News reported that obligations could create uncertainty for technology companies and internet users.

  • 03

    Distinguish the online safety bill from the proposed AI framework, which Ticker News says also uses broad definitions of risk and harm.

  • 04

    For AI governance planning, track the announced proposals for security controls, serious model-risk management and independent evaluations.

What the digital duty would change

The bill would replace the existing industry code framework and Basic Online Safety Expectations with a digital duty of care.[3] The Department of Infrastructure says online services would have to take reasonable steps to keep Australian users safe.[2] SBS reports that the scope includes social media, search engines, apps, games and AI chatbots.[4] The reforms would also let users choose whether algorithm-recommended content appears in their feeds and empower the eSafety Commissioner to issue removal notices for nudify apps and websites.[2] Providers that fail to comply could face penalties of up to A$109.2 million if the legislation passes.[2]

Why the proposed duties concern critics

Bushnell argued that the duty would apply to almost every online service and cover both harms defined in legislation and a wider range of potential risks.[1] Ticker News reported that these obligations could create uncertainty for technology companies and internet users.[1] Bushnell also said users might not know how a proposed law affects services behind the scenes.[1] The report said compliance costs could make it harder for smaller competitors to enter the market.[1] Bushnell described the government’s productivity messaging as “window dressing.”[1]

Concerns about enforcement and moderation

At a Senate inquiry, Liberal senator Sarah Henderson argued that categories such as “hostile attitudes towards women” could restrict legitimate debate.[4] TikTok representative Jessika Lofstedt warned that platforms could over-moderate contentious material to avoid penalties, and said the bill replaced certainty with ambiguity and subjectivity.[4] The government presents a different case: Communications Minister Anika Wells said the bill would establish basic safety standards and require companies to address risks in their products.[4] The bill’s third reading was agreed to on 28 October 2026.[3]

A separate proposed AI framework

The concerns about the digital duty sit alongside a separate proposed AI regulatory approach.[1][6] Reclaim The Net reported that Prime Minister Anthony Albanese announced planned Australian AI Standards legislation on 10 November 2026, with mandatory AI safety and security protections.[6] The same report said the government released 18 recommendations from a review of a June AI incident and accepted all of them; the incident involved access to a government database, but no personal data was stolen.[6] Announced requirements for a proposed AI regulator include security controls, management of serious model risks and independent model evaluations.[6]

Professionals responsible for online products or AI systems should distinguish the proposed digital duty from the separate AI framework and track how each develops. The reported concerns make it useful to assess where broad risk definitions, compliance demands or moderation choices could affect your work.

Why it matters
For teams

AI training for your team

kju.ai is built for teams: daily sessions, team analytics, and rollout support for your organization.

How this developed

  1. 15 November 2026

    Australian AI and online safety proposals draw concerns over broad duties

Sources

AI fluency, one session a day, built for your work.