A photograph that theoretically portrayed Adonis Georgiadis recently took the Internet round. Something that was weekend and the news was not dense, which was supposed to be about a minister at tender moments with unknown elements lady in a restaurant, the rumors raged. And Adonis Georgiadis and the newspaper whose logo appeared in the photograph were forced to intervene, to deny the rumor that the photograph was about him and ultimately to prove that it was a report coming from Serbia and involving a politician of that country.
The incident would have been an isolated one, if in government circles the concern about the phenomena of deepfakes had not been respected recently, as we enter the final line for the election. The examples from neighbouring countries are fresh: Italian Prime Minister Georgia Meloni appeared in photographs posing in underwear, while in Cyprus with the case "Emily Thompson“ A storm broke out over the financing scenario for close associates of President Nikos Christodoulides. In Athens, the threat is considered to be absolutely real and in the face of elections the degree of alarm is further increasing.
Warning
Mr.Mastrogeorgiou does not speak at any chance, as the risk has been highlighted in closed discussions at the government level, which involve the NSA, the National Cybersecurity Authority, the Ministry of Digital Governance and EL.A.'s Electronic Crime DEA that "scan" the Internet for the prospect of threats. institutional interventions are already on the table. Justice Minister George Florides announced a legislative initiative on deepfake videos, with the government studying European models - such as that of Denmark - to criminalise the malicious use of image and voice of citizens through AI. At the same time, measures for greater transparency on digital platforms and limitation of anonymous online manipulation are being considered.
However, the sound "kakanaki" comes not only from government officials, but also from the annual report of the NSP for 2025 published in mid-March. The EPA integrates deepfakes, and more generally misinformation through Artificial Intelligence, in the context of the "hybrid threats" that it considers will increasingly affect national security and democratic stability.
"At international level Artificial Intelligence is already used by malicious actors to automate attacks, create malicious software, undermine education algorithms and falsify data or inputs. Also, the ability to produce deepfakes, i.e. false images, videos or sounds, makes it possible to create material that can mislead the general public or even disrupt the institutional function of states. In many Western countries there are already attempts to use it maliciously to influence electoral processes or promote false narratives and undermine confidence in state institutions," the report points out.
Misinformation
The NSA says that "disinformation campaigns with false news" can now be organized in ways that are difficult to identify, while hybrid attacks can come from both state and non-state actors. That is also the crucial bet: whether a private individual can attempt to intervene or these hybrid attacks are traced back to the bouquet of options that have third countries in order to influence political developments.
The report's "spirit" is that AI instruments - including deepfakes - can be used as weapons of psychological enterprise, destabilization and manipulation of information, especially in times of intense political polarization. "Adonis had this time a great margin to intervene and correct at the level of public debate an obvious lie. But what if we have a similar or even more serious incident 48 hours before the election? “ He wonders with meaning the experienced government executive involved in the discussions taking place at government level, reflecting, therefore, the degree of reflection on the complex and full - time digital environment in which we go.