Overview
In recent years, AI has moved into news organisations – key gatekeepers to the public sphere and central players in our information ecosystem. In many cases, ‘traditional forms’ of AI have been used in both news production and distribution for years without readers or journalists really noticing. This changed with the public launch of ChatGPT in November 2022 by US start-up OpenAI and the subsequent release of further Large Language Models and the rise of what has been called ‘Generative AI’. Many news organisations are excited by this form of AI, due to the possibility of easily creating authentic, multi-modal content (e.g., texts or images) and the manipulation of language more broadly speaking.
While it is too early to say if some of the loftier promises and expectations around the use of generative AI in and for news will be borne out by reality, issues around their effects have started to shift into focus. These include worries about the accuracy of output, copyright and liability, privacy, and bias, effects on trust, to name just a few. Yet, one area that has largely gone unaddressed so far is the effect of generative AI on publishers’ business models and audience reach – and how the increasing use of generative AI might further entrench existing power asymmetries between the news industry and the technology sector or create new ones.
While past work on the relationship between news organizations and the technology sector has focused on the perspective of publishers, the opposite view has not been so well addressed. What AI start-ups such as OpenAI and Anthropic or platform businesses such as Google, Meta and Microsoft want from news organisations, how they see the role of journalism in the age of AI, and how they see their role and responsibilities in this relationship is largely unknown.
This project seeks to fill that gap by asking the following questions:
- What are the motives and interests of both platform companies and AI start-ups around AI, news, and journalism?
- How does the AI sector exercise its power over AI infrastructure and services, how do publishers respond, and what are the implications for the production and circulation of news?
Photo:
Image generated using Dall-E