We saw nations like Germany and France and now Italy put their reservations on the side and that has paved the way for other states in the EU to vote in favor of the last compromise in the much talked about AI Act.
While activists are celebrating the ordeal, many critics have become super wary of the exceptions on this side of the spectrum.
The final vote arose during the committee that featured members from every nation. There were deputy reps to the European Union. When we saw the ambassadors rubber-stamping the ordeal, it was a big endeavor that first came into the public eye last year in December.
This means the AI Act was green-lighted to be voted across the EU Council. It held the official decision-making powers and that’s when the decision first came across the European Parliament.
Once the majority comes into play, we shall see the AI Act get published in the EU’s Official Journal and transform into a law in about 20 days.
The entire AI Act would put a ban across all systems that have been outlined as having serious risks like those seen in different US states or social scoring formats found in places like China. These are used to classify individuals depending on the behavior at stake.
💻 Historical moment at #Coreper today!
— Niels Timmermans - (he/him) (@timmermansniels) February 2, 2024
🇪🇺 After 3 years of work and efforts by 7 presidencies, Ambassadors @EUCouncil just approved the final compromise text on the #AIAct.
👏 First of a kind in the world! https://t.co/vloepAYrYE pic.twitter.com/LZroC2lB6R
This is a draft legislation that is deemed to be one of the latest comprehensive AI laws out there today. It sets out new limits in terms of the high-risk AI and which systems get the chance to impact voters or harm an individual’s overall health.
To be more specific on this front, we’ve got plenty of generative AI models like ChatGPT that are subject to be targeted here. This would force the AI Act to make content generated through such means attain labels so people are more informed about what’s going on.
Last but not least, the bill would prohibit all such models from rolling out summaries regarding copyrighted data.
For a while, we’ve been seeing the impact of this AI Act be in constant threat thanks to countries like France and Germany as well as Italy. Meanwhile, the opposition put into place in terms of the control linked to foundation models was also a topic of debate in this regard.
For months, all EU member states were called out and told not to undermine this particular AI law by saying yes to all claims regarding AI Act adoption. They felt adopting the law would cause heavy regulation that would halt creativity and innovation.
The EU regulator also felt that it was not right to leave the regulatory burden across smaller actors and that’s when we witnessed a few countries take a step back. Both Germany as well as Italy added how they were on the rise to prevent double regulation across phases like implementation. There were many reps who added how Europe needs to design competitive AI models while not overburdening firms having high-risk obligations.
The European Parliament also said it was unacceptable to leave all the regulatory burden on smaller actors, and the trio finally relented.
Germany and Italy sent signals that would seek to avoid double regulation in the implementation phase, and France’s representatives said Europe still needs to build competitive AI models and avoid overburdening companies with high-risk obligations.
This is a major step taken in the EU and one that critics feel was a long time coming.
Read next: A New Study Shows that AI Can Help a Lot in Resolving Cases About Copyright Infringement
by Dr. Hura Anwar via Digital Information World
No comments:
Post a Comment