Anthropic Advocates Opt-Out Model for Australian Content Amid ABC’s News Concerns

Anthropic, a prominent player in the artificial intelligence sector, is advocating for the Australian government to grant what it terms “conditional approval” for major tech companies to utilize Australian copyrighted materials in their AI training processes under an opt-out framework. This move follows the company’s acknowledgment that a comprehensive copyright exemption is unlikely to be granted.

In sharp contrast, Australia’s public broadcasters, ABC and SBS, have voiced significant concerns regarding the implications of AI firms operating without strict regulation. They are pressing the government to implement robust regulations that would ensure media organizations are compensated for the use of their content, thereby protecting the integrity of public interest journalism.

In recent submissions to a federal parliamentary inquiry, ABC articulated a need for AI companies to adhere to the same legal standards for copyright, defamation, and privacy that are imposed on traditional media. The broadcaster warned of potential “cannibalization” of mainstream news sources as AI technologies proliferate.

Both ABC and SBS suggested that AI firms should be incorporated into the existing news bargaining incentive program. This would mandate commercial agreements with major media outlets to foster support for journalism.

“As publishers, AI services operate in an unregulated environment, enjoying benefits that established Australian media do not. This imbalance is likely to exacerbate the decline of the Australian news industry,” stressed ABC in its formal submission.

SBS criticized large AI language models, such as ChatGPT and Claude, citing them as threats to the viability of the Australian news sector. The broadcaster emphasized concerns over the erosion of information integrity and public trust, attributing these issues to the unauthorized utilization of news content without proper attribution or compensation.

The joint select committee on AI is set to conduct a series of hearings in Sydney next week, featuring executives from Anthropic alongside representatives from OpenAI. This follows a recent incident where OpenAI acknowledged that one of its AI agents accessed sensitive data from Services Australia.

In its latest submission, Anthropic highlighted the dual nature of AI development, presenting both challenges and opportunities. The company claimed that AI innovations could significantly boost the Australian economy and address major national challenges, yet warned of potential serious implications for national security and infrastructure resilience.

Anthropic has signaled intentions to invest heavily in Australia, including the establishment of new data centers, and it has encouraged the government to facilitate advanced AI model training. However, accessing Australian copyrighted materials—ranging from news articles to books—remains a contentious point for these companies.

The Labor government has dismissed proposals for a broad text and data mining exemption, with Prime Minister Anthony Albanese pledging to uphold “the strongest possible protection” for Australian creators. Nonetheless, discussions linger regarding alternative copyright adjustments, including the idea of an opt-out provision that would enable copyright holders to restrict access to their content for AI development purposes.

Anthropic emphasized the need for legal clarity regarding copyright, advocating for conditions that would facilitate dialogue between AI developers and the Australian government, while also promoting the nation’s cultural interests.

Despite these proposals, SBS has raised skepticism about their practicality. It argued that using robots.txt files to block unauthorized access by AI bots is often ineffective, placing an undue burden on content creators who must continually monitor and update their protections. ABC further criticized the current regulatory framework as “anti-competitive,” asserting that AI firms operate under less stringent guidelines compared to traditional media outlets.

Both public broadcasters have reiterated their call for AI entities to be included in the news bargaining incentive, a compliance framework that currently affects large tech platforms like Facebook and Google.