EU Commission proposes criminalizing AI-powered child abuse


The European Commission has announced that it is considering criminalizing artificial intelligence-generated imagery and deepfakes portraying child sexual abuse (CSA) to update laws in response to technological advancements.

The proposal includes introducing a new criminal offense for live-streaming child sexual abuse. It also seeks to criminalize the possession and exchange of “pedophile manuals.” The proposal also talks about going against artificial intelligence (AI) chatbots used for child abuse.

The European Commission aims to update rules from 2011 with the proposal, suggesting changes to mandatory reporting of offenses. These measures are part of a broader package aimed at preventing child sexual abuse, enhancing online risk awareness, facilitating crime reporting for victims, and providing support, including the right to financial compensation.

The Commission’s impact assessment conducted before the proposal highlighted increased online presence and “latest technological developments” as factors creating new opportunities for child sexual abuse. To address online safety risks for children, the proposal encourages member states to invest in “awareness raising.”

There are worries that the technology is misused to create child sexual abuse material (CSAM), and concerns have risen that this artificial content could complicate the identification of real victims for law enforcement authorities. The widespread growth of generative AI has provided lawmakers with a clear motivation to review the regulations.

Related: EU member states ‘endorse political agreement’ to advance AI Act

However, the final form of the proposals will be decided by the European Union (EU)’s co-legislators in the Parliament and Council. According to the Commission, upon reaching an agreement on amending the current directive on combating CSA, it would come into effect 20 days after publication in the Official Journal of the EU.

According to the ordinary legislative procedure, when the European Commission submits a proposal, the European Parliament (EP) and Council of Ministers may approve or amend it. If disagreements persist, both can propose amendments. Negotiations follow, and upon conclusion, both institutions can vote for or against the final version.

In May 2022, the Commission proposed additional legislation focusing on digital services’ obligation to employ automated technologies for detecting and reporting CSAM and identifying grooming activity targeting children. According to the Commission, the current proposal complements the 2022 regulation proposal.

However, there has been a significant increase in awareness of the dangers of deepfakes and AI-generated images in the under two years since the private-message-scanning proposal was introduced.

Magazine: AI has killed the industry’: EasyTranslate boss on adapting to change