MEP Patrick Breyer (Germany, Pirate Party), one of the few representatives fighting for preserving rights online rather than against them, has posted a summary about the EU Parliament's assessment of the proposed "Chat Control" legislation. In short, the "Chat Control" proposal violates basic human rights:
The experts made clear that an "increase in the number of reported contents does not necessarily lead to a corresponding increase in investigations and prosecutions leading to better protection of children. As long as the capacity of law enforcement agencies is limited to its current size, an increase in reports will make effective prosecution of depictions of abuse more difficult."
In addition, the study finds: "It is undisputed that children need to be protected from becoming victims of child abuse and depictions of abuse online... but they also need to be able to enjoy the protection of fundamental rights as a basis for their development and transition into adulthood." It warns: „With regards to adult users with no malicious intentions, chilling effects are likely to occur."
There is an obfuscated link at the bottom of his post to the study, Proposal for a regulation laying down the rules to prevent and combat child sexual abuse: Complementary Impact Assessment. He also has older overview of the problems with the proposed legislation at his blog, too.
(Score: 2, Interesting) by sonamchauhan on Sunday April 16, @01:07AM (6 children)
With minor in a park, there are always adults monitoring interaction between their minors and strangers who interact with them
What about creating functionality where intelligent agents like chatgpt monitor conversations between minor children and strangers and flag those at any hint of child sexual abuse?
(Score: 4, Insightful) by maxwell demon on Sunday April 16, @05:02AM (5 children)
Given that ChatGPT often invents “facts”, I don't think it is a good idea to rely on it for determining crime.
The Tao of math: The numbers you can count are not the real numbers.
(Score: 0) by Anonymous Coward on Sunday April 16, @06:11AM
He used an LLM to try to take the jobs of AI specializing in facial recognition, voice transcription, etc.
(Score: 2) by sonamchauhan on Sunday April 16, @08:36AM (3 children)
The (human) controller will need to make up his or her mind by themself. The GPT would only be a "something suspicious" alarm
Parents are always told to "be aware of what your kids are doing online"
But how can they do so effectively? This tech can balance out the scales.
(Score: 3, Insightful) by pe1rxq on Sunday April 16, @09:38AM (2 children)
Because history has tought us those human controllers are idiots themselves? A mob of parents will efficiently burn any suspect alive whenever the AI shouts 'witch'.
(Score: 1) by pTamok on Sunday April 16, @12:25PM
Humans. Tchah!
BBC: Whispering game - The paediatrician confused with a paedophile has become a cautionary tale against hysteria over sex offenders. But the details have become confused, even down to whether it was a male or female doctor. What really happened? [bbc.co.uk]
(Score: 2) by sonamchauhan on Sunday April 16, @10:23PM
Of course humans are foolish and rapacious.
But how can parents "burn alive" a user that is anonymous? At best they can report this to higher-up authorities if they want any action besides terminating a chat. Surely, there'll be some sense somewhere along the chain
We already have nervous parents sweetly mumbling "OK, lets go sweetie" to their kids, when they figure the stranger conversing in the park is a bit too strange.
This would be the same. An AI guardian would be the same. It could even be a passive participant in the chat, saying hello smiling benignly until it decides action is needed.