AI is great.
But not only ...
Artificial intelligence (AI) makes many things easier and more convenient. But AI can also pose a threat to journalism and democracy. That is why it is important to define clear rules for the use of AI in ORF.
ChatGPT hit the ground running in October 2022 - it quickly became clear that this was something really big. AI as a grab bag for all creatives. And for the lazy. And for everyone in between.
The development is rapid: first a chat tool with amazing answers. Then MidJourney, which generates fantastic images from simple sentences. And recently, videos can also be created from simple sentences - so-called prompts. A new, fantastic world is opening up.
However, fantasy and journalism must not collide. The problem: AI works with probabilities. How likely is it that one word will be followed by the next? If the training data is insufficient or incorrect, this can lead to distortions or even hallucinations. The result may be 80 percent correct, the rest may be made-up nonsense. Or the AI is deliberately used to generate fake news. A false sentence put into the mouth of a political competitor can influence elections.
What can be done, will be done. There is always someone. But we at ORF must not do everything just because it is possible. The credibility of our work must always come first. There must be no compromises. That's why we need clear rules and guidelines for the use of AI in ORF.
Translations, spell checking, data visualization, image search, research tools, speech-to-text, ... of course. This opens up new possibilities for the quality and quantity of our work. But there must also be clear limits: What we put on air must be genuine and not artificially generated.
We must not let anyone say something in their own voice that has never been said. We must not "invent" images for reporting just because they look more spectacular or because there is no real footage. Info presenters must not say things on air with the help of AI that they have never said in reality.
The basis for our guidelines must be:
AI is a tool and not a substitute for journalistic work.
The responsibility always lies with the editorial team, keyword "human-in-the-loop". AI content must therefore not be published automatically - without editorial control.
Credibility is our most valuable asset - we must not jeopardize it with "invented reality" in the news section.
Transparency towards our audience: If AI has created a significant part of a journalistic product, then we must be transparent about this.
Our audience must be able to rely on this: ORF news is genuine, verified, independent and therefore credible. AI is not necessarily just a threat to journalism, but can also be a tool for better journalism. .
Most recently, FH-Prof. Dr. Reinhard Christl also dealt with this question as part of a study; Prof. Dr. Uwe Hasebrink, PD Dr. Jan-Hinrik Schmidt and Dr. Stephan Dreyer from the Hans Bredow Institute also published their analyses on the digital future and public service algorithms on zukunft.ORF.at. The next Public Value Study will also be dedicated to this topic and publish an analysis of various public media projects on the use of AI
ORF's Innovation Officer, Stefan Kollinger, is driving the technology forward.
The world of journalism and media is constantly changing. One of the main drivers of this change is the rapid development of artificial intelligence. As Austria's largest media company, ORF has recognized the signs of the times and is intensively exploring the opportunities and challenges of AI.
With the "AI DAY" initiative, ORF is creating a space for the future. Colleagues from different departments, internal and external experts come together here to discuss the potential of AI in the media sector. The events attract well-known companies that inspire with their presentations on generative AI, synthetic data and diverse applications in administration, content production and distribution.
At the heart of these initiatives is the ORF AI LAB - a creative and technological incubator that serves as a hub for AI activities. Under the motto "Join Forces", this lab initiates innovative projects that show how AI could change the world of media. One tangible product of this culture of innovation is AiDitor, an in-house development by ORF that functions as an independent web application. This tool bundles various AI services and offers users a "playground" to experience the possibilities of AI in an uncomplicated way in their everyday work. The AiDitor enables the creation of texts, images, audio, social media posts, translations and transcriptions as well as a chat function. The aim is to simplify work, secure ORF's technological leadership in the exciting field of AI and support the further expansion of multimedia workflows and products.
However, the use of AI in journalism goes far beyond ORF. Media companies around the world are using AI technologies to automate processes, personalize content, develop new formats and much more. Algorithms support journalists with research, data analysis and the creation of texts. Chatbots take over communication with readers and audiences. AI-supported recommendation systems ensure that users can find content that is relevant to them and discover much more. The possibilities of AI in journalism are fascinating, but they also bring challenges. How can we ensure that AI-generated content meets journalistic quality standards? How do we prevent algorithms from reproducing stereotypes or spreading misinformation? And how do we shape collaboration at the human-machine interface in such a way that the creativity and expertise of journalists and media professionals can generally continue to develop productively and positively?
ORF is addressing these questions with openness and a willingness to shape the future. The AI DAYs are more than just a series of events, they are a commitment to the future in which technology and content go hand in hand. By exchanging ideas with experts, promoting internal innovations and critically reflecting on current developments, ORF is making an important contribution to responsibly shaping the journalism of tomorrow.
Most recently, FH Prof. Dr. Reinhard Christl also addressed the issue in a study; Prof. Dr. Uwe Hasebrink, PD Dr. Jan-Hinrik Schmidt and Dr. Stephan Dreyer from the Hans Bredow Institute also published their analyses on the digital future and public service algorithms on zukunft.ORF.at. The next Public Value Study will also be dedicated to this topic and will publish an analysis of various public service media projects on the use of AI.
