The European Union has taken its first legal step to ban artificial intelligence practices that produce images of child sexual abuse. The governments of the member states of the European Union on Friday, March 13, proposed that the practice be officially included as a prohibited activity in the bloc's main AI regulation.
The proposal would add a ban on AI technology capable of producing child sexual abuse material into the EU AI Act legal framework, the world's most comprehensive artificial intelligence rules adopted by the European Union two years ago.
This step comes amid growing global concerns about explicit sexual content generated by AI systems, including intimate deepfakes that can be created with generative technology. Regulators in various countries assess that the technology has the potential to expand digital exploitation of children and make it difficult for law enforcement.
Pressure on technology companies has also increased after reports emerged about sexual content generated by the AI chatbot owned by Elon Musk's company. The chatbot, Grok, was developed by Musk's AI company xAI and integrated with the X social media platform.
A number of technology regulators in Europe are currently investigating potential misuse of the technology. Regulatory authorities in the UK, Ireland, and Spain along with EU regulators are investigating a case of sexual deepfakes allegedly produced by Grok.
The new proposed ban still needs to get approval from the European Parliament before it can be officially adopted. Members of parliament are scheduled to vote on a similar proposal in the near future.
If both parties agree on the same approach, the next step is negotiations with the European Commission to include the changes in the existing AI legal framework.
The debate is expected to be fierce. The European Commission previously proposed to loosen some parts of the AI rules to encourage technological innovation. The move was supported by a number of major technology companies and business actors, but was criticized by civil society groups and privacy protection organizations who considered the policy to be too accommodating to the interests of large technology companies.
Analysts estimate that the legislative and negotiation process regarding this rule change could take about a year before the new rules are actually implemented.
This phenomenon demonstrates the paradox of modern generative technology. The same machine that is capable of writing poetry, making music, or helping scientific research can also be used to create highly destructive visual manipulations. In computer science terms, algorithms have no morals - they only maximize patterns from data. Morality always comes from the human who writes the rules.
The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)
