On 8 October 2026, Denmark took a step of relevance far beyond its own borders. It introduced a bill that would make it unlawful to distribute realistic digital replicas of a person’s face or voice without their consent, and would require platforms to remove the unlawful material swiftly. The protection extends to all citizens, as well as to performers whose work is digitally reproduced. Though not yet law, this initiative must not remain a Danish exception. Every state, Greece included, should establish comparable protection, adapted to its own legal order, before it is forced to react after the damage has been done.
Bring it closer to home for a moment. You turn on your phone and see your own face saying something you despise. You hear your voice, recognize your gestures, and yet it is not you. People who know you will ask whether you said it; anyone who does not will already have decided. A false report distorts an event, but a deepfake is worse, because it also steals the person who seems to verify it. It turns you into the vehicle for a lie and forces you to answer for words that were never yours.
The harm cannot be measured in views alone. It can wound a relationship, damage a person’s professional standing, or expose them to threats. By the time the forgery is exposed, the material is already out there. The correction travels more slowly and with less spectacle than the lie, while the suspicion lingers. The victim is not simply asking for a file to be taken down; they are asking to be given back control over their own identity.
For democracy, the stakes extend beyond individual harm. Accountability presupposes that we know who spoke and who must answer for those words. We can and should question the meaning of a politician’s words, the motive behind them, and their sincerity.
That is the essence of democratic scrutiny. But if we do not know whether the words are actually theirs, we no longer know whom we are supposed to hold to account. Severing a voice from a speaker robs responsibility of its anchor and undermines political judgement.
During an election campaign, a few minutes are all it takes to make a candidate seem to endorse an extreme position or confess to an unlawful act. By the time the denial arrives, the reactions have already taken shape. There is a reverse threat too: a real speaker can dismiss an authentic recording as fake and bid us not to believe our eyes. A fabricated video can be presented as evidence, and a genuine one as a fabrication. No democracy can demand accountability when truth is dictated by whoever is believed first.
The European Union’s Artificial Intelligence Act imposes transparency obligations on certain categories of artificially generated or manipulated content. A notice that a video was created with AI may warn the viewer. But it does not answer the more personal question: who gave someone the right to take my face, copy my voice, and put someone else’s words in my mouth? A label may reduce deception; it does not undo the shock of seeing yourself turned into someone else’s mask. The European framework protects the public’s right to know. The Danish proposal protects the individual’s right to say no.
This protection cannot become a shield for the powerful against satire. Politicians and public figures have to tolerate exaggeration, caricature, and uncomfortable criticism. Consent cannot give them the right to erase whatever exposes them. But the right to criticize does not include the right to place a false confession in someone’s mouth and present it as real. The Danish bill attempts this delicate distinction, exempting satire while narrowing that exemption where disinformation creates a specific, serious risk.
The boundaries will be tested in practice—not in a neutral laboratory, but in the lives of real people. A video can begin as obvious satire and, minutes later, circulate without a caption or context as supposed evidence. A label might appear on the original post and vanish from its copies. Authorities and courts will have to ponder whether a reasonable viewer could have been deceived, what the purpose of the manipulation was, and how serious the potential harm. The cost of getting it wrong, in either direction, will not be theoretical: it will either restrict a free voice or leave a human being defenseless against a lie wearing their own face. Democracy cannot protect truth by silencing satire, nor defend freedom by abandoning people to counterfeit versions of themselves.
For the person being targeted, “we’re looking into it” offers no protection while the video continues to spread. Swift removal is necessary. Yet we cannot hand the platforms unchecked power to decide what is deception, what is satire, and what must be taken down.
We need clear deadlines, reasoned decisions, independent review, and an immediate avenue of appeal. Speed without safeguards becomes censorship; safeguards without speed leave the victim alone against the lie.
A face copied in Copenhagen may have gone viral in Athens before any national mechanism has even swung into action. The dignity of a European citizen cannot be protected in one country and abandoned in the next, simply because a file has crossed a digital border. Europe needs common minimum safeguards for voice and face, rapid cross-border removal, protection for satire, and meaningful access to justice.
Denmark has opened a debate that will soon concern every democracy. Freedom of expression remains fundamental, but no one should be allowed to appropriate another person’s identity and manufacture speech in their name. When anyone can be made to appear to say anything, the right to remain who you are becomes a precondition of public accountability. And if tomorrow neither our face nor our voice belongs to us, what exactly will remain of our right to be ourselves?
