The European Commission on Tuesday recommended a set of measures to be taken by IT giants that own platforms and online search engines with more than 45 million active users in the EU to reduce systemic risks that could affect the upcoming European Parliament elections in June.

Under the Digital Services Act, designated services with more than 45 million active users in the EU are required to reduce the risks associated with electoral processes while protecting fundamental rights, including the right to freedom of expression.

Measures recommended by the EC before, during and after the elections:

• strengthen their internal processes, including by resourcing internal teams, using available analysis and information on risks specific to the local context and on users’ use of their services to search for and obtain information before, during and after elections to improve their mitigation measures.
• promote official information about electoral processes, implement media education initiatives and adapt referral systems to empower users and reduce the monetization and virality of content that threatens the integrity of electoral processes.
• political advertising must be clearly marked as such in view of the new provision on transparency and targeting of political advertising.
• adopt specific mitigation measures related to generative AI: very large online platforms and very large online search engines whose services may be used to create and/or distribute generative AI content should assess and mitigate specific risks , related to AI, for example clear labeling of AI-generated content (e.g. deep fake)adapting their terms and conditions accordingly and ensuring that they are properly enforced.
• cooperate with national authorities, independent experts and civil society organizations to promote effective information sharing before, during and after elections, and to promote the use of appropriate mitigation measures, including in the area of ​​Foreign Information Manipulation and Foreign Interference (FIMI) , disinformation and cyber security.
• take specific measures, including an incident response mechanism, during the election period to reduce the impact of incidents that may have a significant impact on election results or voter turnout.
• assess the effectiveness of measures through post-election reviews. Very large online platforms and very large online search engines should publish non-confidential versions of these post-election review documents, giving the public an opportunity to receive feedback on the mitigation measures taken.
• cooperate with the European Digital Media Observatory (EDMO) working group on the 2024 European elections.

What will happen if the IT giants do not follow these recommendations

The European Commission states that these recommendations represent “the best practices for mitigating the risks associated with electoral processes at the moment”, but their implementation depends on the specifics of each online platform and search engine.

• “Therefore, very large online platforms and search engines that do not comply with these guidelines must demonstrate to the Commission that the measures taken are equally effective in mitigating the risks.
• If it receives information that calls into question the appropriateness of such measures, the Commission may request additional information or initiate formal proceedings under the Digital Services Act,” the European Commission said in a statement.

Minister of digitalization: I do not want the world to panic that we are going to ban TikTok / The new mechanism will remove illegal content in a maximum of 60 minutes

Bohdan Ivan (PSD), Minister of Digitalization, reacted on Tuesday after PSD and PNL leaders told HotNews.ro that the TikTok platform must be urgently regulated in this election year. “I don’t want the world to panic because we’re going to ban TikTok or any other app from tomorrow, but we’re talking about very clear rules that will govern when we’re talking about deepfaking, misinforming people,” Ivan said. . TV B1.

The minister noted that it is a matter of weeks until there will be a clear regulation of social networks, so that messages that incite hatred, hidden hatred of the state, and that violate democratic rules are eliminated. The rules will apply to all social networks.

• “We have two levels. The technical level that, after the analysis carried out by our ministry with the public authorities on cyber security, we came out with a clear recommendation from the Operational Council on Cyber ​​Security, according to which everything that means a service device from the public administration in Romania is strongly recommended not to install this program.
• The second aspect, which is more related to political marketing, is very clearly regulated by the Law on Artificial Intelligence, which was adopted by the European Parliament, the Romanian Strategy on Artificial Intelligence, which we are going to adopt in the Government and which has been publicly discussed for more than six months, and the Law on Digital Services, a directive at the European level, which is automatically transposed into Romanian law and for which, since December, we have created an enforcement mechanism in a form where everything that means content incites hatred and hatred of a state that comes and violates democratic rules or is created by artificial intelligence and does not adheres to the platforms terms of use – we are talking about the standard – this is filtered much more strictly and automatically ends up being harder to break.
• We have also created a mechanism that will be implemented in a maximum of two to three weeks, by which anything that means such reported content can be removed in no more than 60 minutes,” Bohdan Ivan said.

The minister also explained how this control will be carried out: “There are technical filters, and when you use a certain verbal pattern, certain words, you use a certain behavioral profile, the filters automatically prohibit the upload of that video clip, for example. Regardless of the party, regardless of whether he is a politician or not. We are talking about a single standard that does not benefit anyone.”

Photo source: Awargula / Dreamstime.com