Stuff joins media firms blocking ChatGPT from learning from their … – Stuff.co.nz
Media company Stuff has moved to prevent artificial intelligence engines such as ChatGPT from using its stories to learn how to answer questions and create content.
Chief executive Laura Maxwell said in a statement Stuff had joined the growing group of news organisations that were blocking ChatGPT owner OpenAI from using software to trawl its websites and “scrape” information.
It has also updated its websites to make clear that AI companies are not allowed to use its content to train their software models.
Other media organisations that have shut off their content to OpenAI reportedly include CNN, The Guardian, Reuters, the New York Times and Australian broadcaster ABC.
The scraping of websites by software tools is a widespread practice that is needed for a variety of purposes, such as the effective operation of internet search engines, and is surrounded by its own etiquette.
OpenAI made it easier for website owners to implement blocks on its software last month by providing information that helps publishers identify the GPTBot tool that it uses to harvest information.
Maxwell said the scraping of any content from Stuff or its news masthead websites for commercial gain had always been against its policies.
ChatGPT caused waves when it launched in November by demonstrating the capabilities of so-called “generative AI” to a broad audience for the first time.
Reuters reported last week that web traffic to ChatGPT had declined for the third month in a row, to total just over 1.4 billion visits in August, but that there were signs the popularity of the service might now be stabilising.
Maxwell said there were growing concerns that AI would exacerbate the spread of “disinformation and misinformation globally”.
However, the primary goal of media businesses in blocking the GPTBot appears to be to persuade OpenAI to pay licence fees to make use of their information.
There had been a surge of unease from news organisations, artists, writers and other content creators that their work had been harvested without permission and without compensation by companies seeking to build new commercial products through “generative AI”, Maxwell said.
“The news industry must learn from the mistakes of the past, namely what happened in the era of search engines and social media, where global tech giants were able to build businesses of previously unimaginable scale and influence off the back of the original work of others,” she said.
Although commercial agreements had not been widespread, OpenAI had entered into negotiations with some news organisations in the United States to license their content to train ChatGPT, naming Associated Press as an example, she said.
Stuff was looking forward to holding conversations around licensing its content in due course, Maxwell said.
© 2023 Stuff Limited
