Edit2 : I misunderstood the breadth of that bill and you can ignore the rest of my post below.
I suggest reading my last comment on this post that links to the EFF report on the No Fakes Act.
Sorry !
The NO FAKES Act create a broad, new right to restrict AI-generated depictions of a person’s voice and likeness — and in the process expose users, creators, platforms, and developers to lawsuits for everyday speech like memes and parodies.
For the other two laws I completely agree that they are dangerous.
But for the NO FAKES act, in my opinion it’s actually a useful one. AI fakes in my opinion serve no purpose and can very easily be used to spread misinformation. And as far as I can tell as long as you meme without using generative AI you can caricature and mock your politics as long as you want.
I’m surprised a community that is very much anti-AI seems to oppose this NO FAKES act ?
Bear in mind that I’m commenting from the perspective of an European so our conception of freedom of speech and its limitations is culturally different.
I’m certain AI fakes will be used nefariously to manipulate voters and if the cost is AI memes, so be it.
It’s too bad this article conflates privacy invading laws with a law restricting the use of generative AI which is pretty much completely unregulated as of now ?
Edit : We have an EU regulation for this type of AI fakes coming into effect in August 2026 for anyone interested :
The EU AI Act is the world’s first comprehensive law on Artificial Intelligence and applies across all EU member states. It doesn’t ban deepfakes but imposes strict transparency obligations to ensure citizens are not unknowingly deceived. It is set to take effect on August 2, 2026.
Without dogging too far, it seems like the law is broadly worded enough to open the door for all sorts of SLAPP-type takedowns, a bit like how DMCA is weaponized against people that don’t have armies of lawyers.
Also, the other source (not the bill itself, mind you, so might be wrong) says “digitally generated”, not “AI generated”, which could be stretched to apply to any image manipulation, like cropping.
Then of course there’s the question of reliably differentiating between AI and non-AI. Which basically means whoever has the biggest legal cannon to fire at the other guy wins.
Unless there’s also a flawless system that can decide what is and isn’t AI generated, it will just be used to shut down anything anti government, while keeping pro government slop up. The way EU did it sounds better, but could still be abused in the same manner - just take down anything the government doesn’t like and say it was because they didn’t label it as AI generated, have fun going through the court system trying to prove it wasn’t.
AI fakes will always be an issue. Nothing will ever prevent someone from creating AI content and passing it as real. Especially with the complete lack of regulations on AI model makers.
At the same time AI fakes are inherently dangerous for democracies in the sense that anybody could anonymously spread misleading fakes all over social network to disrupt the democratic process and it would be very difficult to fight.
I don’t know what solutions we can find for that. I guess we will see in the EU if our approach is effective or not. Or if it is abused to censor speech. Considering our own lawmakers are also digitally illiterate I wouldn’t be surprised if there is loopholes to exploit.
Edit2 : I misunderstood the breadth of that bill and you can ignore the rest of my post below. I suggest reading my last comment on this post that links to the EFF report on the No Fakes Act.
Sorry !
For the other two laws I completely agree that they are dangerous.
But for the NO FAKES act, in my opinion it’s actually a useful one. AI fakes in my opinion serve no purpose and can very easily be used to spread misinformation. And as far as I can tell as long as you meme without using generative AI you can caricature and mock your politics as long as you want.
I’m surprised a community that is very much anti-AI seems to oppose this NO FAKES act ?
Bear in mind that I’m commenting from the perspective of an European so our conception of freedom of speech and its limitations is culturally different.
I’m certain AI fakes will be used nefariously to manipulate voters and if the cost is AI memes, so be it.
It’s too bad this article conflates privacy invading laws with a law restricting the use of generative AI which is pretty much completely unregulated as of now ?
Edit : We have an EU regulation for this type of AI fakes coming into effect in August 2026 for anyone interested :
The EU AI Act is the world’s first comprehensive law on Artificial Intelligence and applies across all EU member states. It doesn’t ban deepfakes but imposes strict transparency obligations to ensure citizens are not unknowingly deceived. It is set to take effect on August 2, 2026.
Without dogging too far, it seems like the law is broadly worded enough to open the door for all sorts of SLAPP-type takedowns, a bit like how DMCA is weaponized against people that don’t have armies of lawyers.
Also, the other source (not the bill itself, mind you, so might be wrong) says “digitally generated”, not “AI generated”, which could be stretched to apply to any image manipulation, like cropping.
Then of course there’s the question of reliably differentiating between AI and non-AI. Which basically means whoever has the biggest legal cannon to fire at the other guy wins.
I tried checking the actual bill and it seems you are correct and its target is not only limited to AI generated fakes but everything digital.
I have now found a summary of the EFF about the NO FAKES act and I’m most likely misunderstanding the breadth of that bill :
https://www.eff.org/files/2024/09/12/2024.11_no_fakes_one_pager.pdf
I almost blindly trust the EFF on these types of questions, they have been an invaluable ressource even in Europe to protect citizens.
I suppose the European approach is probably safer here since it doesn’t limit speech but mandates transparency on AI fakes.
Unless there’s also a flawless system that can decide what is and isn’t AI generated, it will just be used to shut down anything anti government, while keeping pro government slop up. The way EU did it sounds better, but could still be abused in the same manner - just take down anything the government doesn’t like and say it was because they didn’t label it as AI generated, have fun going through the court system trying to prove it wasn’t.
I have since edited my original comment.
AI fakes will always be an issue. Nothing will ever prevent someone from creating AI content and passing it as real. Especially with the complete lack of regulations on AI model makers. At the same time AI fakes are inherently dangerous for democracies in the sense that anybody could anonymously spread misleading fakes all over social network to disrupt the democratic process and it would be very difficult to fight.
I don’t know what solutions we can find for that. I guess we will see in the EU if our approach is effective or not. Or if it is abused to censor speech. Considering our own lawmakers are also digitally illiterate I wouldn’t be surprised if there is loopholes to exploit.