Oštro, Center for investigative journalism in the Adriatic region, nurtures investigative and data journalism and champions the »right to know« as one of the fundamental human rights, as well as contributes to the cultivation of future generations of journalists. Oštro’s newsroom is committed to upholding the professional standards as set out in the Code of Slovene Association of Journalists, the Charter of Munich, IFCN code of principles, EFCSN Code of Standards and Oštro’s Code of Conduct.
One of Oštro’s fundamental principles is to investigate matters of public interest and verify the authenticity of information published in the media. Our editorial team upholds rigorous professional journalistic standards. We do not publish any information that has not been thoroughly verified.
Below is an explanation regarding the methodology and objectives of content labeling within the framework of the Rapid Response System (RRS).
As mentioned above, our newsroom follows strict professional standards, including our own internal code of conduct. Because of these commitments, we do not publish or flag information unless it has been thoroughly verified.
We flagged only one post during our monitoring of the 2026 Slovenian national elections – the only case of great importance to the public debate where we could say with complete confidence that the information was not credible. When we had indications that content may not be credible but our analysis did not produce clear results, we chose not to flag it.
We flagged the post on 23 March 2026 for content which we had previously verified using the appropriate tools (Hive) and two separate analyses conducted by Slovenian experts in the field of artificial intelligence. We flagged the content on Facebook (two posts) and on Instagram Threads (one post). We also flagged these posts via Meta’s internal tool on 18 March 2026.
Another reason we have issued a single flag is that, throughout the last election period, our weekly monitoring of elections-related statements dispersed during live public debates, via social media posts, general online discourse and media reporting has largely revealed individual claims that fell into two categories: misinformation – such as incorrect statistical data or misinterpretations thereof – and isolated cases of disinformation.
Our assessment is that a significant portion of the 2026 election campaign has taken place on social media platforms. Many candidates have chosen to appear as guests on podcasts hosted by various Slovenian influencers (with no journalistic background), rather than to participate in debates moderated by professional journalists.
A few days before the election, there was an unexpected development. Covert recordings of conversations between current and former officials of the leading party, the Freedom Movement, began circulating on Social Media. The individuals featured in these videos included a Member of Parliament, a former party Secretary-General, and a former Minister of Justice. Some of them, including the Freedom Movement, claimed that all or parts of the recordings were AI-manipulated. Once published, the recordings triggered widespread debate about their authenticity and the possibility that they were released with the intention of influencing the election.
A selection of these materials was forwarded to an independent specialist body, that operates under the auspices of the international Misinformation Combat Alliance, Deepfakes Analysis Unit – DAU, dedicated to deepfake analysis for further examination. Drawing on these expert evaluations, our team developed comprehensive fact-checking content. Nevertheless, we did not flag any content related to these recordings because, on Meta platforms, we detected only news articles discussing them – not standalone posts containing the videos.
A few days before the March 22 election day, the public first learned about the alleged foreign (Israeli) interference in the Slovenian elections at a press conference, where a journalist from the weekly newspaper Mladina and two representatives from the Slovenian NGO sector presented their report on the matter.
Members of Oštro’s team attended this press conference. The main piece of evidence presented by Mladina was given to their journalist by an anonymous source to whom we consequently had no access.
Soon after the press conference, the Slovenian intelligence agency (SOVA) confirmed that Israeli nationals indeed visited Slovenia on the dates referenced in the report. Additionally, Janez Janša, the president of the SDS party which hosted the Israelis in question, confirmed to Slovenian media that he had previously met with one of the individuals mentioned in the report.
SOVA provided further evidence on the 26th of March at the meeting of the National Security Council. It has informed the Slovenian police and the public prosecutor’s office of suspected criminal activity regarding the operations of a foreign private intelligence agency and its contacts with Slovenian entities.
Although further information has since begun to surface, the circumstances surrounding the Black Cube affair were too vague that – in line with our methodology and Oštro’s Code of Conduct – we decided to refrain from flagging any related content.
We want to emphasize again that Oštro adheres to a strict policy requiring every lead or source of information to be verified with primary sources and the highest attainable level of certainty. That has not been possible in this case.
Under the Slovenian Election and Referendum Campaign Act, the campaign period for parliamentary elections must end 48 hours before election day. This year, the campaign concluded on Friday, 20 March 2026, at midnight, and the ban remained in effect until polling closed on Sunday, 22 March 2026, at 7 p.m.
In May 2022, the Constitutional Court of Slovenia upheld this rule and said that the period of so‑called “election silence” applies to everyone: not only political parties, but also private individuals, legal entities, and the media. During this period, it is prohibited to disseminate or publish information directly related to the elections, with limited exceptions, such as real‑time reporting of election results or reporting on violations of the silence period.
Given these restrictions, and based on our monitoring, we did not detect any content during the “election silence” period that could have been reported to the RRS.