The ‘Double Life’ Of Artificial Intelligence: From The EU AI Act To Cyprus’ Response

Artificial intelligence is often presented as either a blessing or a curse. In reality, it has the potential to become both. While AI is transforming modern life in remarkable ways, it is also creating entirely new forms of abuse capable of violating privacy, dignity and fundamental human rights.

One of the most alarming examples is the rise of “deep fakes”, which are realistic but fabricated images, videos or audio recordings generated through artificial intelligence. What once appeared to be harmless digital experimentation has now evolved into a serious social and legal concern. Explicit content can be created within minutes using ordinary photographs taken, for example, from social media accounts, even without the victim ever knowing.

We are living in an era of constant technological exposure. Some individuals willingly share aspects of their personal lives online, and that is entirely acceptable. What is not acceptable is that, at the same time, people may be exposed in ways they never chose themselves, whether they are active users of social media or not, without their consent, knowledge or control. Artificial intelligence has intensified this reality dramatically. A simple photograph can now be manipulated into false sexual content and circulated instantly, causing humiliation, emotional distress and reputational damage that may be impossible to reverse.

At the same time, the dangers extend far beyond personal attacks or violations of privacy. AI-generated content is increasingly being used in cases of financial fraud, identity theft and online scams, where fabricated audio or video recordings are capable of imitating real individuals with alarming accuracy. There have already been international incidents involving fake voice recordings of company executives used to deceive employees into transferring substantial sums of money.

The political sphere is equally vulnerable. Artificial intelligence can now generate entirely fabricated statements, speeches or videos showing politicians and public figures appearing to say or do things that never actually occurred. Such material can easily be used as a tool of propaganda, manipulation or public intimidation, particularly during election periods or moments of political tension. In an age where information spreads instantly and public opinion is shaped within seconds, the damage caused by false but convincing digital content can be enormous.

The growing concern is therefore not only whether people can distinguish truth from fiction, but whether society itself may gradually lose trust in what it sees and hears.

The European Union had already recognised the growing dangers associated with artificial intelligence through the EU AI Act (Regulation (EU) 2024/1689), which entered into force in 2024 and introduced a common regulatory framework for AI systems across Member States. Among other obligations, the legislation imposed transparency requirements regarding AI-generated or manipulated content, including deepfakes. Where an AI system is used to generate or substantially alter image, audio or video content that resembles real persons, objects, places or events and could be mistaken for authentic material, the user must clearly disclose that the content has been artificially created or manipulated. This disclosure must be made in a visible and distinguishable manner, including appropriate labelling of the AI output and an indication of its artificial origin. At the same time, the Regulation clarifies that such transparency requirements should not be interpreted as restricting freedom of expression or artistic freedom, especially where the content forms part of creative, satirical, fictional or artistic works, provided that appropriate safeguards for third-party rights are respected and the disclosure does not undermine the normal use, enjoyment or dissemination of the work.

More recently, the Council and the European Parliament reached a further agreement aimed at simplifying certain aspects of the framework while at the same time strengthening safeguards against harmful AI practices, including the generation of non-consensual explicit deepfake material.

Cyprus has also taken significant legislative action. In March 2026, the House of Representatives approved legislation aimed at addressing the non consensual creation and dissemination of deep fakes. The relevant law strengthens the protection afforded to individuals in relation to the unauthorised use of their image, voice and other personal characteristics, recognising such attributes as integral components of personal identity. It provides for both civil remedies and criminal sanctions against those responsible. The amendment reflects the growing understanding that personal identity deserves legal protection not only in the physical world, but equally in the digital environment.

At the same time, Cyprus has already witnessed the real-life consequences of AI misuse. In a recent judgment, the Limassol Juvenile Court imposed, among others, custodial sentences on two secondary school pupils (12 and 9 months respectively), suspended for three years, following their conviction for using AI applications to generate and distribute fake pornographic images images of underage female classmates and a teacher. The fabricated images were then circulated through online messaging groups, causing serious psychological and emotional harm to the victims.

Legally, the case signals an important development: AI-generated material is not treated as being outside the scope of existing criminal offences. Even though the images were not “real photographs”, the harm produced is real and sufficient to engage criminal liability under existing frameworks dealing with child sexual abuse material, privacy violations and related offences.

From a broader policy perspective, the case highlights three emerging risks. First, the normalisation of AI tools among minors without corresponding awareness of legal consequences. Second, the speed at which reputational and psychological harm can be caused once content is circulated online. Third, the difficulty of containment once such material enters closed digital networks.

A similar case came before the District Court of the British Bases in Dhekelia in October 2024, involving a 14-year-old student who circulated manipulated images of his female classmates at a secondary school in the Larnaca district. The incident caused significant disruption, prompting intervention by the school administration and the British Bases Police. The student had admitted the offence and the judge did not impose any sentence on the 14-year-old. Instead, the court issued a three-year conditional caution, with the possibility of the matter being brought back before the court if the minor reoffends. This would include any repetition of the same conduct, namely the creation and distribution of child sexual abuse material.

And yet, despite these dangers, artificial intelligence should not be demonised. It is an extraordinarily useful tool which, when used responsibly, ethically and with moderation, can assist in solving problems and everyday difficulties. Artificial intelligence, when used properly, may improve efficiency, facilitate communication, and support education and research. It also has the “superpower” of saving one of the most valuable resources of human life: time.

Τhe danger begins when technology ceases to be a tool and starts becoming a substitute for human judgment, thought and interaction. Artificial intelligence should not paralyse human thinking, nor become an addiction or the mechanism that thinks and decides in place of us. It should remain exactly what it was intended to be; a means that assists human intelligence, not one that replaces it. Technology should be “food” for the mind, not replace the mind itself.

For this reason, the answer should not be fear, but responsible use. Society should neither blindly accept or follow artificial intelligence nor reject it entirely. Instead, it must learn how to use it correctly, ethically and within clear boundaries.

And where individuals fail to respect those boundaries, the responsibility inevitably passes to the State, through its legislative, executive and judicial authorities, and to society as a whole. The law must intervene to protect human dignity, privacy and consent through effective regulation, enforcement mechanisms and, where necessary, meaningful punishment for those who misuse technology to harm others. Because ultimately, no technological advancement can truly be called “progress” if it comes at the expense of humanity itself.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.