The European Commission is proposing that major online platforms and search engines label deepfakes, detect foreign manipulation networks early and account for their actions after every election, under draft guidelines opened for public consultation on Tuesday until November 23, EFE reports.
According to the EU executive, the text responds to ‘a rapidly evolving threat landscape, with increasingly sophisticated risks of manipulation’, as well as to the European Council’s call for an urgent review of the current recommendations.
The draft recommends new measures for platforms to implement before, during and after each round of elections. These measures should be operational between one and six months before the vote, strengthened in the final month and maintained for at least one month after the elections.
Ahead of elections, platforms should identify and dismantle networks engaged in coordinated inauthentic behaviour at an early stage, as a defence against foreign information manipulation and interference (FIMI).
To curb fake accounts, platforms should verify user registrations, limit the number of accounts per person and display the country from which each account is managed. They should also test their recommendation algorithms before every national election to ensure they cannot be manipulated, and log any changes made during the six months preceding the vote. During election campaigns, they should identify suspicious trends before they go viral and clearly label deepfakes, with those labels remaining in place when the content is shared.
Details, HERE