GENERATION „KI“

Journalism International
Dec 17, 2023
3 min read

Updated: Nov 28, 2024

The expansive use of artificial intelligence

Radio, television, online print, podcasts: today the world is full of multi- and cross-media media. According to statistics on the user group of selected media in Germany in the years 2014-2022, in cooperation with Seven.One Media, (publication October 2022), books, newspapers and magazines were the most popular after broadcasting (90%; n=2,453). in printed form, but also as an e-edition or podcasts as the most popular medium in 2022.

Fig. 2: Quantitative telephone interviews conducted by AI throughout Germany (Source: Statista)

The age group from 14 to 69 years old was surveyed by telephone throughout Germany in the spring of 2014-2022. This quantitative method, explicitly the Ex Post Facto Design: Panel, is also a computer-assisted telephone survey (CATI).

But this raises the next research question:

“How public is the use of algorithmic, computer-generated intelligence to date?”

“This text was written by an algorithm”

According to the draft regulation laying down harmonized rules for artificial intelligence (AI-VO) presented by the EU Commission, Article 52 ABS. 1 & 3 (KI-VO), there is a transparency requirement for synthetic media content that was created or modified by AI, in particular deepfakes. Accordingly, an obligation to disclose should be imposed when using an AI system if image, audio, video or text content is created or manipulated that cannot or can hardly be distinguished from authentic content. Made externally are s.g.n. High-risk AI systems that, according to the European Parliament, “pose a high risk to the health and safety or fundamental rights of natural persons”. (Europäisches Parlament, 2023)

This includes, among other things, the biometric identification and categorization of natural persons in law enforcement or in the administration of migration, asylum and border controls.

The transparency obligation therefore requires the name of the natural or legal person to be stated when it is created or manipulated. This means a clear, visible label that informs you that content was generated algorithmically.

International use of AI in media

The “Quakebot” software from the US Los Angeles Times can be classified as a pioneer of computer-generated news. On March 17, 2014, in response to an alert from the U.S. Geological Survey, which follows an earthquake of a certain magnitude, sends a message to the in-house newsroom within seconds, which can then immediately inform the public of the important breaking news of an earthquake.

But in other countries, such as Germany, algorithmic support is already being used in the media. This includes working with augmented reality (AR), which integrates digital elements into a live TV broadcast to expand the environment. These days, especially in live football broadcasts, due to the inclusion of line and circle symbols.

Conclusion:

Not only should a uniform regulation for public use in the media be created legally, but also morally, with regard to the journalistic norm, in particular the ethical press code, there must be standardized labeling when using an AI system. According to the press code, which is not legally established, journalists or freelance PR professionals are subject to the protection of human dignity or freedom of expression, so they must be aware of their responsibility towards the public and their obligation to the reputation of the press (Press Council). A journalist should be held accountable for their own work or receive recognition and an increasing reputation. Which also means that when an AI is used in a designated manner, it should be assessed differently or an AI should only be used as a controlled tool in order to limit equal transparency on the one hand and also limit fake news.

Literature references:

Europäisches Parlament (2023). KI-Gesetz: erste Regulierung der künstlichen Intelligenz. https://www.europarl.europa.eu/news/de/headlines/society/20230601STO93804/ki-gesetz-erste-regulierung-der-kunstlichen-intelligenz

Seven.One Media (2022). Weitester Nutzerkreis (Nutzung mindestens selten) ausgewählter Medien in Deutschland in den Jahren 2014 bis 2022. Media Activity Guide 2022, 34

Oremus, W. (17. März 2014). The First News Report on the L.A. Earthquake Was Written by a Robot. Slate. https://slate.com/technology/2014/03/quakebot-los-angeles-times-robot-journalist-writes-article-on-la-earthquake.html