ZDF sees opportunities, but also risks, in the use of generative AI. It is therefore open and at the same time critical of the topic.
Generative AI tools are intended to support, but not replace, the work of the editors.
The editorial departments are responsible for the use of generative AI to support programming, planning and communicative processes in ZDF. The use is subject to the quality and program guidelines.
Generative AI tools can be used in an editorial and journalistic context. The two-source principle applies, whereby generative AI itself does not count as a source.
Content (text, image, audio/video) created with the support of generative AI is subject to the principles of journalistic due diligence. Content, origin, and truthfulness must be checked.
Content created with the support of generative AI is reviewed, approved, and published by editors. An exception is live subtitling for broadcasts such as sporting events.
Generative AI tools must not be enriched with sensitive data (e.g. personal data, thirdparty data, company secrets). Exceptions are possible after prior examination by editors with the participation of the respective responsible departments.
Content created with the support of generative AI is identified as such for the users whenever demanded by editorial and journalistic principles.
ZDF deals transparently with its own developments in the field of generative AI.
