[ad_1]
The European Parliament on Wednesday voted to adopt the European Union (EU) Artificial Intelligence (AI) Act, enacting comprehensive legislation with lessons for governments around the world.
The regulations were agreed by negotiators in December and approved by MEPs with 523 votes in favor, 46 against, and 49 abstentions.
“We have finally passed the world’s first binding law on artificial intelligence to reduce risks, create opportunities, fight discrimination and bring transparency,” said Brand Benifay, Member of the European Parliament, on Tuesday. mentioned in the main discussion.
At the federal level in the United States, authorities have already taken some steps in this area. The Biden administration announced a voluntary AI Bill of Rights in October 2022, but some experts believe this is only the first of several steps needed to effectively regulate AI. He claimed that there was no. But last October, the administration went further, announcing an executive order setting new standards for AI security and privacy. There are several bills in Congress addressing AI, including the Algorithmic Responsibility Act and the AI Research, Innovation, and Accountability Act. And at the state and local level, regulations are rapidly evolving.
But the passage of the EU AI law raises questions about what U.S. government leaders can learn from it.
New America, a nonprofit, nonpartisan think tank, hosted a panel discussion in February titled “EU AI Law: Lessons for U.S. Policymakers,” where several experts discussed the EU’s approach to AI governance. .
Prem M. Trivedi, policy director at the Open Technology Institute (OTI), said during a panel discussion that the EU AI law is expected to have a global impact.
“As such, U.S. policymakers must not only consider the text and requirements of the law, but also ensure that national efforts in the U.S. and around the world reach the goal of broad global regulatory harmonization among like-minded countries. We also need to consider how we can promote this,” Trivedi said.
What is prohibited under EU AI law?
Gabriele Mazzini, AI Team Leader for the European Commission’s AI Law and lead author of the AI Law proposal, shared her expertise in a panel discussion.
Mazzini pointed out that the EU AI law does not seek to regulate AI technology itself. Instead, we regulate specific use cases through a risk-based approach.
More specifically, there are certain AI systems that are prohibited under EU AI law as an “unacceptable risk”, such as systems that perform social scoring or compile facial recognition databases.
Laura Lázaro Cabrera, legal advisor and director of the Equity and Data Program at the EU Center for Democracy and Technology, said during a panel discussion that there are values to highlight when considering the application of this law to other jurisdictions. He said that one of the positive aspects of an EU AI law is that: That human rights are central.
He also pointed out that the law considers the prohibition of certain actions to be positive. However, she argued that there are exceptions to these prohibitions in this regard.
She further explained that while Parliament has called for a total ban on remote biometric identification, the EU AI law contains exceptions to this.
“So while it is great that these prohibitions are in place, there are some caveats to some of them that may undermine the prohibitions in the future,” she said, adding: He said it was not yet known how far the ban would extend.
What should our leaders take away?
Mazzini pointed out that the goal has always been to maintain the understanding that activities carried out within the EU are not self-isolated. He said: “Given the global nature of technology, we intentionally wanted to align internationally, at least on certain key points and elements, such as the definition of AI.”
David Mueller, senior policy analyst at OTI, said during a panel discussion that defining international standards is an area where U.S. and EU leaders can work together.
Mueller said EU AI law will become the basis for U.S. regulation. While the U.S. may make changes, companies are likely already aiming to comply with the requirements of EU AI law, so the regulatory discussion should start there, he said. The main benefit of the bill, according to Mueller, is that the US and his EU will now speak the same language on issues of his AI policy.
He recommended that the United States move forward by first addressing privacy.
“In terms of lessons learned…the first thing the United States can and should do is pass a comprehensive federal privacy law,” Mueller said.
Kai Zenner, Member of Parliament for Axel Voss and Digital Policy Adviser, said during the panel discussion that there is currently a lot of international cooperation on risk assessment and that cooperation is a positive force. Still, Zenner said more time is needed to see how the law works in practice, how different interpretations of the law might arise, and whether further regulatory action is needed. Stated.
“With every new law, there’s going to be a transition period where we have to really find each other first,” Zenner said.
[ad_2]
Source link
