Anthropic has urged the Albanese government to grant "conditional approval" for AI companies to train models on Australian copyrighted works, proposing an opt-out system that would let copyright holders block their content from training data1,2. The proposal, filed as a submission to a federal parliamentary inquiry, comes after Anthropic conceded it will not secure a blanket copyright exemption in Australia.
The company suggested the government "permit AI model training through a narrow form of conditional approval" and proposed that rights holders could signal refusal through robots.txt files. Anthropic said broad copyright exceptions such as a text-and-data-mining exemption "have not received widespread support and have been ruled out by the Government". The Labor government has separately ruled out such an exemption.
Public broadcasters push back
Australia's public broadcasters took a sharply different position. The ABC demanded in its submission that AI companies be subject to the same copyright, defamation, and privacy rules as media outlets, warning of a "cannibalisation" of mainstream news. The ABC called the current system anti-competitive, arguing that AI-based services are unregulated and receive more favorable treatment than established Australian media.
SBS echoed those concerns, stating that AI-powered large language models such as ChatGPT or Claude are "challenges to the sustainability of the Australian news industry and the erosion of information integrity and public trust through their use of news content without permission, attribution or compensation". SBS also argued that robots.txt mechanisms, the very tool Anthropic proposed as an opt-out mechanism, "are routinely ignored or bypassed". Both broadcasters called for AI companies to be included in Australia's news bargaining incentive, the framework that already compels platforms such as Facebook, Instagram, Google, TikTok, and Snapchat to negotiate with news publishers.
Anthropic's economic and security pitch
Alongside its copyright proposal, Anthropic framed AI development as economically and strategically critical for Australia. The company claimed the technology "could transform the Australian economy and drive a new wave of productivity" and "solve some of the nation's biggest challenges". Anthropic also warned of "serious implications for Australia's national security, the resilience of its critical infrastructure and the safety of its people" if frontier model training is not supported. The company said it plans to invest billions in Australia, including large new data centers, and said it was open to supporting Australian creators.
Prime Minister Anthony Albanese has promised "the strongest possible protection for Australian creatives". The joint select committee on AI will hold a series of hearings in Sydney next week.
ANALYSIS Anthropic's pivot to a conditional opt-out model, after the blanket exemption path closed, represents a middle-ground bid that still places the burden of action on rights holders rather than on AI companies. SBS's claim that robots.txt is routinely circumvented directly undercuts the technical mechanism Anthropic proposed, setting up a core factual dispute the parliamentary inquiry will need to resolve.