ABC trials Claude AI for journalism in partnership with Anthropic
Australia's ABC will test Anthropic's Claude AI in newsrooms to convert radio programs into articles, marking a shift in the broadcaster's stance on generative AI use.
Australia's ABC has reversed its previous caution toward generative artificial intelligence, striking a deal with US tech company Anthropic to trial Claude AI in its news production. The trial marks a significant shift in how the national broadcaster approaches automation in journalism.
The trial scope
For now, the use of Claude is limited. The ABC will focus initially on converting radio programs into written articles—a workflow change designed to free up journalists for more demanding work such as investigations. The broadcaster has signalled, however, that it may expand the AI's role to other editorial tasks as the trial progresses. The ABC plans to hire specialists to support staff through the adoption process.
Why newsrooms are exploring AI
Journalists and news producers have a long history of embracing new technologies to streamline their workflows. Automated systems and "robo-writing" tools that convert data into news stories have been in use for over a decade. A 2016 study found that readers even rated computer-written articles as "more credible and higher in journalistic expertise."
The goal is to free up time for staff to engage in core journalistic work, such as investigations, while expanding the ABC's production capabilities.
The current wave of generative AI tools differs from earlier automation, however. Because chatbots can write in human-like language, they raise new concerns among journalists about the future of the profession itself.
The risks ahead
Journalists are more wary of generative AI than they were of earlier automated tools, according to reporting on the trial. Part of their anxiety stems from the technology's capacity to produce plausible but false information—a risk that has already caught out professionals in law, medicine, and other fields relying on AI outputs.
For newsrooms, this risk is particularly acute. Journalism already faces a crisis of public mistrust in Australia and globally. Introducing generative AI into news production without robust oversight could amplify concerns about credibility and accuracy, or displace journalists at a time when the profession is already struggling with sustainability.
The ABC's decision also reflects a broader tension in media. While technological innovations like social media and analytics promised to enhance journalism, they have sometimes undermined editorial autonomy by placing algorithmic judgments above newsroom decisions. Generative AI introduces questions about who ultimately controls news production and what safeguards protect quality.
Public reception
Australians have shown distinctive distrust of AI tools more broadly. How the public will respond to the ABC's trial remains unclear, particularly as the trial expands beyond the initial radio-to-article conversion work.
The broadcaster's willingness to test generative AI sits within a longer tradition of newsrooms adopting cutting-edge technologies. Whether this trial becomes a model for sustainable AI integration in journalism—or a cautionary tale—will depend on how the ABC addresses ethics, quality control, and the role of human editorial judgment.