OpenAI signs agreement with News Corp to access its content and optimize ChatGPT
-
Sulaiman
- . May 24, 2024
- . Update: May 24, 2024 8:51 pm
A new agreement for OpenAI. Wednesday 22nd May, OpenAI, the maker of ChatGPT, and News Corp, the international media conglomerate, have signed an agreement that allows OpenAI to utilise and learn from News Corp’s content.
Other titles in the American group covered by the agreement include the conservative daily New York Post, the tabloid The Sun, The Times and The Sunday Times in the UK, as well as The Australian.
In practice, this means that when a user asks a question to ChatGPT, the results may be informed by previous reporting in News Corp-affiliated media, including Australian newspapers such as The Australian and The Daily Telegraph.
OpenAI has licensed News Corp’s content because AI is a predatory beast: It needs data to learn from and generate useful outputs in return.
Accusation of copyright infringement
At the end of December 2023, The New York Times filed a lawsuit against OpenAI and Microsoft. Eight local US newspapers also filed a lawsuit at the end of April. For a few months now, several authors, artists and news sites have been accusing OpenAI and its competitors of copyright infringement in the generative artificial intelligence (AI) race, which requires mountains of data.
Keep Reading
The ChatGPT creator is in the process of signing content licensing agreements with media outlets – including Le Monde, the Associated Press, Germany’s Axel Springer Group and Spain’s Presa Media Group – to enrich their models.
After actress Scarlett Johansson accused OpenAI of deliberately copying her voice for its new voice assistant without her knowledge, the announcement of the deal with News Corp. Sam Altman, CEO of OpenAI, apologised and announced that the voice, named Sky, would be discontinued.
At the end, generative AI doesn’t always get it right (and often gets it wrong), so treat the output with a healthy level of caution and compare results with those from trusted sources before using AI-generated content to make decisions.
