An Election Operations Center with experts to identify potential online threats, strict control over deepfakes and content with artificial intelligence which also includes sanctions.
These are the commitments of Meta, parent company of Facebook and Instagram, in view of the European elections, explained in a post by Marco Pancini, Head of EU Affairs of the company.
Particular attention is paid to contents generated by artificial intelligence for which Meta has already signed a commitment to label them with common standards together with other big names in technology.
One option for deepfakes (audio, video and images) is to mark them as "altered" and reduce their visibility in the news feed.
Furthermore, photorealistic images created with tools from Meta but also from Google, OpenAI, Microsoft, Adobe, Midjourney and Shutterstock will be indicated clearly and if necessary more visible and with content that explains the context.
Users who share AI-generated audio and video will need to add a label to avoid sanctions, such as deletion of the content or suspension of the account.
The same obligation applies to advertisers who use generative artificial intelligence for political content.
Regarding political ads, Meta recalls the platform's transparency process on advertisers thanks to the 'Paid by' notice to show who is behind each ad.
"Between July and December 2023 - says Pancini - we removed 430,000 adverts across the EU because they did not include a disclaimer".
"We believe it is important that people know when the photorealistic content they see has been created with artificial intelligence - he concludes -. This work goes beyond that of a single company and will require an enormous effort from the industry in the sector , governments and civil society".
Reproduction reserved © Copyright ANSA