From black box to glass box journalism

AI in the selection and verification of (relevant, diverse and reliable) information?

For the journalistic research, editing and verification process, more and more self-learning AI tools are entering the market. DataMinr, for example, deploys such algorithmic tools to sift through millions of Tweets to help journalists track the latest trends. While AI tools and search engines are efficient, they are far from objective. These tools rank information according to user popularity. Moreover, information that ranks higher on a search page is usually seen as more relevant and reliable. This gives search engines a crucial role in the information process.

That algorithms are not neutral calls for research into the use of automated tools in journalism. In what ways are journalists influenced by AI in the selection and verification of (relevant, diverse and reliable) information? We examine four aspects:

  • To what extent are AI tools used in the daily journalistic research process?
  • To what extent are journalists aware of the use and operation of such AI tools?
  • How do AI tools affect the diversity and reliability of journalistic information?
  • How can the operation of AI in the journalistic research process be made more transparent?

Researchers

Yael de Haan, Nele Goutier, Eric van den Berg

From black box to glass box journalism