The Parliament confirmed the thoughts of Wednesday where it added how the goal was to regulate generative AI to keep citizens safe at all times. Moreover, a whopping 523 votes were received by members who voted in favor of this new act while just 46 voted against it.
It is believed that this new law that governs AI use took a staggering two years to come into shape and therefore is designed to ensure that humans come before AI at all times.
The Parliament mentioned how all practices that it felt were unfit in regards to AI governance would no longer be acceptable in this part of the world while citizens' rights would be the highest priority at all times.
There would also be a new setup for a new AI office in the region to assist firms in complying with a host of regulations before coming into play.
The EU Commission’s head also gave replies to reports on this front by adding how the new AI Act will work towards assisting the EU’s workforce and give rise to a plan where users can blindly trust AI not only in Europe but other regions around the globe.
Passing of this new law also means that after getting finalized, systems would be categorized into one of the four leading risk domains. It would be controlled in the same way. Any AI tools that it feels aren’t acceptable would now be subjected to a serious ban including biometric categorization and AI tool poilcing.
Tools that cover domains such as education, immigration, justice, safety, employment, law enforcement, and beyond could come under a new category called high-risk. But those that carry minimal to no risk would be able to skip the entire oversight completely.
All apps that are considered high-risk wouldn’t be given the green signal from the general public unless they have the necessary paperwork that allows them to function safely and after meeting high-quality standards.
Users coming across AI tools such as chatbots would now feel safer as those tech giants giving rise to such offerings would now disclose the matter to the EU and how they are interacting with technology and not real human beings. This means they’re no longer going to be kept in the darkness.
If and when they feel they’re at threat, they have the right to back out as the reality is printed in front of them and no one will force them to do anything.
The Parliament also mentioned how the main focus for AI systems is complying with copyright laws in this area and also rolling out summaries in detail about any data used for the likes of training and beyond.
Users can protect work online and ensure it’s not copied or made use of for the likes of training large language models for example. So if the user to whom this data belongs says they don’t want it to be offered to others, then it would be off-limits to all.
After this act transforms into law, the EU says the impact can be major for those hosting generative AI models that were trained using a heap of content from the web that’s copyrighted. But it’s still not clear what stance AI firms have on this front and what the actual impact of such laws would be from their end.
Remember, the law will take another three years to come into play completely in the EU while limitations detailed in the new document say it would come into play as early as six months but AI models can take up to one year to follow so there’s a long time coming.
Whatever the case may be, one thing is for sure. It was a long time coming and users are happy that something is being done to safeguard them against the growing woes of the AI world.
Read next: TikTok Users Stage Uproar As App One Step Closer To Getting Banned After Bill Passed In US House
by Dr. Hura Anwar via Digital Information World
No comments:
Post a Comment