AI in Journalism: A Balance Between Creativity and Safety

A recent Dutch column examined the role of artificial intelligence (AI) in journalism. The author wondered whether systems like ChatGPT and Google Gemini are capable of generating authentic columns, or if they mainly remain vague and detached. This discussion also has repercussions for editorial teams and training programs, especially as it relates to European regulations that may influence the use of AI in the media.

The systems from OpenAI, Google, and Anthropic are often configured with so-called safety filters, ensuring that sensitive subjects or neutral texts are the norm. This can dull the sharp edges of creative writing, which is problematic for opinion and style pieces. Beautiful and nuanced writing styles may be lost in the attempt to stay safe, making AI possibly not the best solution in this context.

Editorial teams in the Netherlands are now exploring ways to utilize AI without replacing human journalists. They primarily use AI for tasks such as transcriptions, summaries, and translations, while the involvement of the journalist remains crucial for opinions and columns. All of this occurs within the framework of internal guidelines that ensure transparency and control over the use of AI. Editorial teams are expected to document and make visible the use of AI in their processes for their readers.

With the arrival of the European AI regulation, which will come into effect in 2025, AI providers must also offer more transparency about their methods and data. These laws aim to clarify the responsibilities of both users and providers of AI. Given the importance of recognizability and personal voice in journalistic content, the question remains whether AI is the right partner in the creative process or whether it serves as a limitation to personal expression, viewed through the lens of European regulation and copyright.

Read the full article from AI Insider.