Lithuanian MP proposes ban on deceptive AI-generated content

Social Democrat Birutė Vėsaitė has proposed banning the distribution of artificial intelligence-generated content that threatens democracy or national security or violates individual rights, LRT reported.

Vėsaitė has registered amendments to Lithuania’s Law on the Provision of Information that would establish a new category of prohibited information. If approved by the Seimas, Lithuania’s parliament, the ban would cover AI-generated or substantially altered content that creates a false impression about real people, their identities, statements or actions.

The proposed ban would apply to content that could unlawfully influence democratic processes, threaten national security or significantly violate people’s rights and legitimate interests.

Vėsaitė said the amendments would reduce opportunities to misuse AI to create misleading content, strengthen public resilience to disinformation and other forms of information manipulation, and provide clearer legal rules for protecting society.

The explanatory note accompanying the proposal says rapidly developing AI technologies can create or substantially alter video, audio, text and other digital content in a highly realistic way. Such technologies can be used to spread disinformation, manipulate public opinion, influence democratic processes, violate individual rights and threaten national security, it says.

The note also refers to a June report by the Council of Europe Parliamentary Assembly’s Committee on Political Affairs and Democracy, which said AI can be used to manipulate elections and other democratic processes, spread disinformation, impersonate people and reduce public trust in democratic institutions.

Under current law, Lithuania bans the publication of disinformation, war propaganda, incitement to hatred, pornographic material and information that violates a person’s honour and dignity or the presumption of innocence. However, it does not explicitly prohibit the distribution of AI-generated or substantially altered content that creates a false impression about real people.

Vėsaitė said such content can sometimes be assessed under general rules concerning disinformation, defamation, the protection of honour and dignity or other legal violations, but that no specific regulation currently exists.