Australian media may require Meta to pay for news content to train AI

Mariia Shalabaieva/Unsplash

Australian news publishers are considering seeking compensation from Meta for using their content to train generative artificial intelligence systems, researchers say.

Here's What We Know

According to The Guardian Australia, Meta has not commented on whether the "publicly available online sources" on which its Llama 2 model was trained included news publications. However, the company could rely on the principle of "fair use" under US copyright law.

There have been precedents in the past for AI developers being sued by publishers like The New York Times, accusing them of illegally using articles to train models. OpenAI has managed to get one such case dismissed.

Professors Monica Attard and Michael Davis believe Australia's code of negotiation with the news media could be a tool to force Meta and other tech giants to pay media companies to use their content for machine learning purposes.

"Because news archives are such a rich, reliable data base for AI training, they have extra value as a training source for AI companies and platforms. And in terms of the known risks to the information environment, its clearly important that AI is trained on high quality data sources" they emphasised.

In December, Australian Attorney General Mark Dreyfus announced the creation of a task force on copyright and AI. Different parties have expressed different positions, ranging from full legalisation to requiring licensing of AI content.

Including companies like Meta in the negotiating code would allow publishers to demand compensation for the use of their content in training AI models or block access to their news archives.

Source: The Guardian