Media give themselves rules for dealing with AI

Artificial intelligence poses new questions and problems for media companies. How trustworthy are the results generated by AI? How can content be labeled? Some media have already issued guidelines, others are still in the middle of the work.

(Iconic image:

The Ringier media group considers a conscious approach to AI to be highly relevant, especially for media companies. Ringier has therefore decided to introduce guidelines for all companies in the group, which is represented in 19 countries. This was announced by the media house on Tuesday ( reported).

Among other things, the results generated by AI tools should always be critically scrutinized. Furthermore, the content shall be labeled. The AI tools and technologies developed, integrated or used by Ringier shall always be fair, impartial and non-discriminatory. Confidential information and trade secrets shall not be entered into AI applications.

Reference to artificial voices at SRF

At Schweizer Radio und Fernsehen SRF, the aim is to achieve maximum security and transparency, as can be seen from the guidelines on the website. For example, when artificially generated or cloned voices are used, explicit reference is to be made to this fact. Furthermore, the company relies on human-centered and controlled AI.

The development of guidelines at the media houses Tamedia, NZZ and CH Media is still in the planning stage. An editorial working group is working on this, CH Media said at the request of Keystone-SDA. In general, the company places the highest value on a responsible approach to AI.

Tamedia and NZZ are on it

The NZZ wrote in response to a query that it was looking into both the content and the technical use of AI and was currently drawing up corresponding guidelines. And Tamedia plans to publish its guidelines shortly, according to the communications manager.

The Keystone-SDA news agency also defined its stance for dealing with AI in April and published it on its website. Among other things, the agency declares when a form of AI has been used in obtaining information or content. Likewise, images are marked where a form of AI has been applied. (SDA)

More articles on the topic