ON
← Back to feed
SloveniaCulture4 days ago

Prohibition of deepfakes with sexual content

The European Parliament has approved amendments to the EU AI Act, banning the use of artificial intelligence systems to create deepfakes containing sexual content or depicting intimate parts of identifiable individuals without their consent. The new rules prohibit AI systems from generating such material, including images, videos, or audio recordings showing explicit activities or child exploitation. Providers of these systems must implement technical safeguards to prevent such creation, and operators using them for this purpose are also restricted. Companies have until December 2nd of this yе

Evropska unija

Spremembe mora potrditi še Svet Evropske unije

16. junij 2026 ob 16.03

Strasbourg - MMC RTV SLO, STA

Al. Ma.

Evropski parlament je potrdil spremembe akta o umetni inteligenci, s čimer je prepovedal uporabo UI-ja za ustvarjanje globokih ponaredkov s spolno vsebino, poenostavil nekatera pravila in zamaknil roke uveljavitve.

Evropski svet se je za prepoved odločil po aferi, ko je orodje umetne inteligence Grok na pobudo uporabnikov ustvarjal seksualizirane vsebine, tudi otrok. Foto: Pixabay/JuliusH

Spremembe akta o umetni inteligenci vključujejo prepoved sistemov umetne inteligence, ki ustvarjajo gradivo o spolnem izkoriščanju otrok ali ustvarjajo slike, videoposnetke in avdioposnetke, ki prikazujejo intimne dele prepoznavne osebe ali spolno eksplicitne dejavnosti brez njenega soglasja.

Oglas

Ponudniki teh sistemov jih ne bodo smeli dajati na trg Evropske unije (EU), razen če bodo opremljeni z ustreznimi tehničnimi zaščitnimi ukrepi za preprečevanje ustvarjanja takšnega gradiva. Prepoved velja tudi za operaterje, ki uvajajo sistem za uporabo v ta namen, so v sporočilu za javnost pojasnili v Evropskem parlamentu.

Podjetja bodo imela za uskladitev svojih sistemov z zakonodajo čas do 2. decembra letos.

Določbo, ki prepoveduje prakse umetne inteligence v zvezi z ustvarjanjem spolnih in intimnih vsebin brez soglasja ter gradiva, ki prikazuje spolno zlorabo otrok, je marca v predlog Evropske komisije za spremembo akta dodal Svet EU-ja.

Za to se je odločil tudi zaradi kritik družbenega omrežja X in njegovega klepetalnega robota z umetno inteligenco Grok zaradi seksualiziranih fotografij žensk in otrok.

Potrjene spremembe zajemajo tudi poenostavitev akta o umetni inteligenci in zamike nekaterih rokov. Tako bi se moralo 2. avgusta letos med drugim začeti uporabljati pravila za visoko tvegane sisteme umetne inteligence. Na podlagi tokrat potrjenih sprememb bodo pravila začela veljati 2. decembra 2027.

Za sisteme umetne inteligence, vgrajene kot varnostni sestavni del in zajete v sektorski zakonodaji EU-ja o varnosti in nadzoru trga, pa bodo nove obveznosti veljale od 2. avgusta 2028.

Prav tako se zamika uveljavitev obveznosti uporabe vodnega žiga na vsebini, ustvarjeni s pomočjo umetne inteligence, do 2. decembra 2026. Gre za tehniko, ki omogoča odkrivanje in sledenje vsebine, ki jo ustvari umetna inteligenca.

Spremembe akta o umetni inteligenci mora zdaj uradno potrditi še Svet EU-ja. Večina določb akta bo začela veljati 2. avgusta letos, so še navedli v Evropskem parlamentu.

Oglas

Read the full article at RTV Slovenija (MMC)
Source document: European Parliament

2 reports

MladinaIndependentCenter4 days ago
Prohibition of deepfakes with sexual content

The European Parliament has approved amendments to the EU AI Act, banning the use of artificial intelligence systems to create deepfakes containing sexual content or depicting intimate parts of identifiable individuals without their consent. The new rules prohibit AI systems from generating such material, including images, videos, or audio recordings showing explicit activities or child exploitation. Providers of these systems must implement technical safeguards to prevent such creation, and operators using them for this purpose are also restricted. Companies have until December 2nd of this yе

Bias read (Center): The article reports on legislative changes related to AI regulations, focusing on the prohibition of deepfakes with sexual content. It presents factual information about the European Parliament's decision, the scope of the ban, and implementation timelines. There is no evident ideological framing, o

Official sources cited

  • government European Parliament
RTV Slovenija (MMC)State / PublicCenter5 days ago
European Parliament Bans Deepfakes with Sexual Content

The European Parliament has approved amendments to the AI Act, banning the use of artificial intelligence systems for creating deepfakes with sexual content. The changes also simplify certain rules and delay implementation deadlines. The ban applies to AI systems that generate material involving child sexual abuse or explicit intimate content without consent. Companies must comply with these regulations by December 2nd of this year.

Bias read (Center): The article presents factual information about legislative changes to the EU AI Act without apparent ideological framing. It focuses on technical and regulatory aspects rather than taking a stance on the issue.

Official sources cited

  • government European Parliament

Go to the primary sources (1)

The official sources this coverage is built on. Read them directly to bypass framing.

  • governmentEuropean Parliament