Skip to main content

Even though AI in journalism is experiencing a real hype thanks to generative AI and tools such as ChatGPT, artificial intelligence is no stranger to the media industry. From deep fake videos to robot journalism – editorial teams have had to deal with the digitalization of their profession and AI for a long time. What is new, however, is the speed at which new innovations are finding their way into our day-to-day work. AI is therefore a much-discussed topic in journalism, bringing with it both advantages and challenges. We take a closer look at what AI in journalism is, what use cases already exist in the industry, how editorial teams can best deal with it and where the journey with artificial intelligence can take us.

Definition: What actually is artificial intelligence?

A single definition for AI is not so easy, as it is a constantly evolving field of research and the definitions change regularly. When asked ‘What is AI?’, many people naturally think of ‘intelligent’. And indeed: AI definitely has to do with intelligent behaviour, but less with real ‘thinking’. Rather, AI is a field of research centred around systems that are designed to observe, learn and act like humans. Machines are being developed that can make independent decisions based on extensive data and therefore behave intelligently. There is ‘strong AI’ as well as ‘weak AI’ – and the latter in particular is intended to support us humans and master individual skills.

One area of AI is becoming particularly important for the media industry: machine learning. Using algorithms, mathematical equations and statistics, AI systems learn to recognise certain patterns from data, use them as a basis for making forecasts and recommendations and also learn to improve themselves.

In particular, the use of natural language processing (NLP), prediction and generation are important tools for processes in media production. Such technologies from the field of machine learning can automate and standardise a variety of processes in editorial offices and newsrooms, from content research to production and publication.

RECOMMENDED TO READ:

You can find out how we use NLP in our editorial software here

RECOMMENDED TO READ: You can find out how we use NLP in our editorial software here

And what is generative AI?

Currently, ‘generative AI’ is the main buzzword. It refers precisely to the algorithms that are used to create, i.e. generate, new content such as text, audio, code or images. ChatGPT is a good example of this: it generates new texts in a kind of chat. Unlike previous chatbots, ChatGPT’s responses are not pre-scripted, but the AI responds in a targeted manner to the questions or prompts it is asked.

AI in journalism is not new: Robot journalism

Language models already existed long before ChatGPT, whose language model was trained with a large number of texts from the internet, made the rounds. AI and algorithms were already capable of processing language or generating texts. In the media industry, the term robot journalism or automated journalism had therefore already become established. Numerous articles on the subject were published years ago. A few examples:

Where did the first robots support journalism? Above all with topics relating to the weather, sport or the stock market. This is because there is a lot of structured information, i.e. data, that the early AI models were able to make use of.

But what has triggered the hype when it comes to AI?

Two developments are new in the use of AI: Ever-improving language models and better usability of the tools. A chat component, such as in ChatGPT, has made it much easier for users to use AI. Prompting, i.e. giving commands to the AI, has become easier and can simply be done in natural language. And thanks to a veritable race for the best language models between OpenAI, Google and the like, the results are also getting better and better. So you could say that the hurdles for AI are disappearing more and more.

AI in practice: Current use cases

As a result of these developments, sports reporting or news on shares are no longer the only fields of application cited: A wide variety of AI software is already in use in many editorial offices and publishing houses. Media professionals are not always aware that there is already a form of machine learning and therefore artificial intelligence behind many small automations.

How is AI already being used in the media industry today?

  • A widespread example of AI in media companies is the use of transcription software. Today, such software is supported by artificial intelligence to quickly transcribe audio and video files. Interviews, recordings or recordings can be converted into text in just a few minutes and used for further processing.
  • AI is also already being used for spelling aids or language translation. A well-known tool here is certainly DeepL from Cologne with its own AI translation technology.
  • Automated text creation or text generation: many small publishers and local journalism in particular are already relying on AI to support the production of articles. Small teams in local journalism can thus cope with the large volume of daily news.
  • AI can already be put to good use in video production. On the one hand, editorial teams can automatically add text, subtitles or a voice-over to moving image content. On the other hand, AI can also automatically create entire clips and snippets from an article. Studio 47, a small regional news channel, for example, uses its own tool to support its full program with a small editorial team.
  • More and more editorial software is already supported by AI and takes the workload off editorial teams when planning and preparing topics. Newsmind Stories, for example, automates many otherwise very repetitive and time-consuming tasks, such as research and monitoring. It also helps to pre-sort stories according to relevance and display similar stories from the competition

Newsmind Stories already works with generative AI

  • Artificial intelligence in journalism also helps with community management. After all, journalists’ tasks don’t just include researching or checking information, editing or publishing: they also interact with their audience and have to moderate. Why AI in this area? AI can pre-filter comments and monitor signal words, for example. Only with modern technology is it possible for editorial teams to gain and maintain an overview of thousands of comments.

Artificial intelligence is also used in archive searches, in linking articles and suitable content, in the automatic setting of print editions, in analytics and predictions for user-centered journalism, in contextual advertising, in the creation of audio spots or even in the generation of entire audio programs and suitable image material. And these are just a few examples that make it clear: AI and machine learning are already deeply rooted in the media industry.

Which tasks could machines possibly perform even better in the future?

Many of the new AI tools such as Midjourney, ChatGPT or DeepL Write are still in their first versions and many companies are in the process of integrating the GPT models from OpenAI or language models from Aleph Alpha into their software and solutions and adapting them to various use cases.

Over time, these applications will become even better and more sophisticated, and users will also become increasingly adept at using the tools. Artificial intelligence could then take on even more tasks, for example against disinformation and fake news or to a greater extent in automatic text creation, in video production, in composing music or even more professionally in image generation. One thing is certain, however: AI solutions in the media industry will not become human-like androids, but will remain a tool of digitalization. They are tools to support journalism.

Advantages of AI: Seize the opportunities

In all of the aforementioned use cases of AI in journalism, the focus is on relieving the burden on humans. AI software helps to make everyday editorial work more productive and contributions better, as AI can take over certain repetitive or time-consuming tasks. This creates time and space for new journalistic formats, in-depth research and more creativity.

We have collected the advantages of using AI in the media industry:

  • By using AI, editorial teams can offer more content and thus contribute to more democracy in reporting – particularly important for local journalism.
  • Broader coverage can reach more readers, which can also lead to more subscriptions.
  • Editorial teams can reduce their costs, as automated processes free up resources and allow them to be used more efficiently.
  • Media companies save further costs by increasing efficiency.
  • AI can improve the quality of reporting, for example in data journalism: AI can recognize patterns in large amounts of data better than humans.
  • Editors can save time, for example by using transcription software or software for monitoring, and use this time for more demanding, creative tasks.
  • More content can generate additional advertising revenue.
  • With AI and data analysis, media can offer more relevant content and provide a better experience for the audience.

If AI is used in such a way that it supports the editor, it strengthens creativity and also the role of the media as a knowledge broker.

The challenge of AI: a threat to the media?

Despite the opportunities offered by AI, there are of course also challenges. Media professionals fear for their jobs, for the quality of journalism or even that they will be replaced in their role as intermediaries. So can AI also be a threat to the media? The answer must first be: Yes. Because every new technology invites abuse.

Responsible use of such technologies is not always guaranteed and the public’s media literacy needs to be further developed. Even today, people find it difficult to recognize fake news, identify reputable sources and keep up with the flood of information and channels. The possibilities of generative AI increase the risk of deep fake videos or false propaganda. Journalism also has a duty of care and must address the challenges:

Quality assurance

When AI systems create content, humans or journalists must check that the content created is of high quality and accuracy. A human should always have the final say and decision-making authority, as AI is prone to errors and is not omniscient. Liability must lie with the editorial team.

Privacy

Editorial teams must ensure that no sensitive data ends up in tools like ChatGPT and similar, and also that there are no violations of personal rights when collecting and analyzing data.

Bias and discrimination

AI systems can be unpredictable and prone to bias. If models are trained on data that unconsciously contain prejudices or discrimination, this can lead to inaccurate reporting. Human oversight is also important in this context.

Whether AI is more of an opportunity or a threat often depends on the human factor. The industry and politics must face these challenges and openly address the concerns. They must ensure that the use of AI in reporting is ethical and responsible – in accordance with the press code and values such as truthfulness and diligence.

Training in the use of AI early on – Therefore, it is even more important for editorial teams to engage with new technologies early and not let themselves be overrun by them. Those who get involved early can also learn early and train in responsible handling. Then AI is a great opportunity for the media industry and can help secure democracy instead of endangering it.

Will AI replace editors?

Can AI really mean a loss of jobs? It’s true that automation takes over tasks that previously had to be done by humans. However, in many cases, these are repetitive and time-consuming tasks. Whether it’s pre-formulating a short tweet from a report, applying metadata, processing data, or transcribing an interview: the actual job of journalists and the creative work look different.

AI can therefore relieve editorial teams of less critical tasks and support core competencies from research to editing, but it does not replace anyone creatively. In the end, editorial teams have more time for the processes that matter: creativity, judgment, responsiveness to gestures and facial expressions in an interview, providing opinions or context – all human qualities for complex situations that no AI can take over quickly.

Only editors can create something new

Furthermore, AI can only generate content based on what already exists: texts from the internet, books, videos, etc., serve as the foundation for generative AI. For this reason as well, AI cannot replace journalists. According to the German Journalists Association (DJV), “journalists have the task of making facts or events public whose knowledge is important for society.” Since they create something new and impart knowledge, they remain essential for a democracy and ensure a “comprehensive range of information in all journalistic media.”

Best practices for successful use of AI

Since artificial intelligence in the media industry not only opens up advantages and opportunities but also brings challenges, it is important to consider how to successfully implement AI. Editorial teams need a plan to accompany the introduction of AI. Those responsible must ask themselves the following questions: How can employees be prepared? Which tools are allowed to be used? Is an AI strategy needed for editorial teams? We have outlined how teams can anchor AI in their newsroom and company culture in our whitepaper (German).

Request for free (German)

Journalism will change – but not for the worse

AI will greatly change not only the way media companies produce news but also how we consume it. However, those who use AI responsibly in journalism mainly open up opportunities for more efficient, relevant, and improved journalism. In such a scenario, AI in the media industry is not a replacement for human labor but rather a tool to support and enhance daily processes. Human judgment, creativity, and critical thinking remain indispensable skills in reporting – and thanks to AI, there will be more time for these as well.

Interested in experiencing efficient editorial processes with AI for yourself?

Schedule a personal talk with Jochen – for free and no strings attached.

By clicking on "Send" I confirm that I have received the privacy policy.

Jochen Schon

More interesting articles that you shouldn’t miss

KI und Mensch im Journalismus
AI

AI in journalism: How the media industry is changing

Even though AI in journalism is experiencing a real hype thanks to generative AI and…
Redaktionsmanagement
Data analytics

How does successful editorial management work?

Given the growing number of new market players and the resulting increase in competitive pressure,…
Themen finden in einem Newsroom
AI

Find content ideas easily: With AI-based monitoring of trends and news

Digitalization has added a constantly growing flood of information, news and sources for media companies…