- Thomson Reuters may license select data to AI companies like OpenAI for training large language models.
- This comes as publishers negotiate content access deals to improve chatbot accuracy.
- Reuters now weighs revenue potential against ethics as interest in AI applications continues booming.
Thomson Reuters
News giant Thomson Reuters is open to licensing select data for large language model development, announced CEO Steve Hasker, joining publishers courting deals with AI creators like OpenAI and Google.
Hasker confirmed early discussions with multiple unnamed providers to supply the company’s vast information repositories.
As publishers, Axel Springer and Apple also reportedly negotiate content access for bolstering chatbot capabilities, data from trusted news companies grows vital for powering accurate generative responses on current events.
Thomson Reuters’ legal datasets and Reuters journalism particularly appeal as high-value training materials.
Publishers negotiate
The ability of Microsoft and partners to responsibly utilize media content remains hotly debated, however, as authors and artists accuse tech giants of copyright violations by scraping copyrighted data secretly.
Just this December, the New York Times sued Microsoft over alleged unlawful generative model training.
AI information ideas
Nonetheless, as consumers increasingly turn to assisted AI for convenient access to complex information, licensing deals grant developers superior quality data for better results.
Reuters now weighs its options, balancing revenue potential with ethics.
Publishers are squeezed by big tech and interest exploding in AI applications.
The moment appears ripe for mutually beneficial partnerships, provided proper consent and attribution get established industrywide.
