MAP Forum

Pegasus wrote: ↑Mon Jul 22, 2024 3:10 pm "Even if the content is entirely artificial and does not represent a real victim, AI-generated child pornography material continues to contribute to the objectification and sexualization of children," it warned.

Fragment wrote: ↑Tue Jul 23, 2024 6:11 am

Pegasus wrote: ↑Mon Jul 22, 2024 3:10 pm "Even if the content is entirely artificial and does not represent a real victim, AI-generated child pornography material continues to contribute to the objectification and sexualization of children," it warned.

Which if they are not real fucking children should not be a fucking crime.

Pegasus wrote: ↑Mon Jul 22, 2024 3:10 pm The use of artificial intelligence (AI) to create child pornography online will grow exponentially and fake images will make it increasingly difficult to identify real victims, the European police agency warned on Monday (22).
Criminals are already using AI tools to commit crimes ranging from online fraud and cyber attacks to the creation of explicit images of minors, Europol warned.
In a 37-page report, the agency claimed that there are already reports of "AI-generated and AI-assisted child sexual abuse material".
"The use of AI, which allows abusers to generate and alter child pornography material, is set to proliferate further in the near future," warned the Hague-based agency.
AI-generated pornographic images "increase the amount of illicit material in circulation and complicate the identification of victims and perpetrators".
More than 30 million children were victims of sexual exploitation and abuse on the internet last year, according to an investigation by the British University of Edinburgh in May.

The crimes range from so-called sextortion, in which criminals demand money from victims to keep images private, to abuse by AI to create deepfakes, videos that are not real, according to the Childlight Global Safety Institute program.
The rise of AI has led to growing fears around the world about the possibility of its malicious use.
"The sheer volume of self-generated sexual material constitutes
"The volume of self-generated sexual material now constitutes a significant and growing part of the child sexual abuse material available online," Europol said.
"Even if the content is entirely artificial and does not represent a real victim, AI-generated child pornography material continues to contribute to the objectification and sexualization of children," it warned.

Jim Burton wrote: ↑Sat Jul 27, 2024 11:44 am I have not seen us address their concern yet:

How is it possible to prove that an image was made by AI and involved no real minors?

This cuts to the heart of their objection to AI generated images. The idea that real youth erotica/PIM will be indistinguishable from this vast body of work, and thus almost impossible to prosecute. The idea "verification" watermarks could be applied to real PIM in a deceptive manner.

All I can think of right now is that programs are made with a predetermined electronic signature to each successive production of each successive download, and the sig is overlaid in some way within the output file. These could be verified by cross-referencing to a database that is encrypted somehow as to make copying that data impossible for forgers.

Without an answer to this, there are very valid concerns about generative AI, at least from a law-enforcement perspective.

Pegasus wrote: ↑Mon Jul 22, 2024 3:10 pm "Even if the content is entirely artificial and does not represent a real victim, AI-generated child pornography material continues to contribute to the objectification and sexualization of children,"

Pegasus wrote: ↑Mon Jul 22, 2024 3:10 pm The crimes range from so-called sextortion, in which criminals demand money from victims to keep images private, to abuse by AI to create deepfakes, videos that are not real, according to the Childlight Global Safety Institute program.