The Internal Market and Civil Liberties Committees of the European Union, they voted overwhelmingly in favor of supporting the AI bill
Although the bill has yet to be fully approved, this week's vote is an important step towards codifying AI laws in Europe and is one of the first of its kind worldwide. The bill previously had a tentative agreement on September and has since undergone a round of negotiations, according to the House announcement.
The latest approved version of the bill, if passed, would ban a number of different uses of artificial intelligence, giving citizens more legal protection and privacy regarding some of the most dystopian potential use cases for AI. Biometric categorization, the “untargeted» gathering images of people from the internet or surveillance footage, the “emotion recognition” with artificial intelligence, social scoring and predictive policing tools (à la “Minority Reportert”) will be made illegal in the EU under the AI Act.
Any type of "manipulation» AI that could be used to exploit human behavior would also be banned.
While the use of artificial intelligence for biometric purposes by the police “forbidden in theoryEU lawmakers have created an exception where police will in some cases be able to use artificial intelligence tools that leverage biometric data, according to the European Parliament announcement. Under the AI Act, police will be able to conduct real-time biometric analysis with artificial intelligence for specific emergencies, such as trying to find missing persons or "preventing a terrorist attack".
Post-event use of biometric tools by authorities is considered “High Risk” under the AI Act and will only be possible if it is directly linked to bringing a criminal to justice. A judge or other legislator would also have to grant express permission for such cases.
The bill also puts safeguards in place for the use of artificial intelligence in "high-risk areas" such as education, health care, banking, law enforcement, immigration and elections. It will also give citizens permission and a clear process to make formal complaints about how AI is used in these areas.
Artificial intelligence tools identified as “general purpose artificial intelligence systems” (GPAI), should comply with applicable EU copyright law and meet transparency requirements. The bill also gives lawmakers the ability to impose stricter requirements on any GPAI models deemed higher risk.
It is unclear to what extent copyright laws could apply to artificial intelligence, and whether images created “in the style” of another artist, for example, could be considered to infringe copyright.
The AI Act would also make it illegal to deepfakes unlabeled, meaning that those who are victimized by AI-manipulated or generated images, recordings or videos could have legal recourse under the Act.
Once the legislation is passed it will take some time to fully take effect. Some parts of the bill will take effect within six months, while other parts of the bill will take two years to 36 months to become active under EU law.
While the EU is among the first to promote legislation regarding the use and operation of artificial intelligence, the US, for its part, is taking these steps of its own.
Last week, the president Joe Biden announced a AI Security Consortium, inspired by a previous executive order that collects more than 200 different companies and institutions such as Microsoft, Nvidia, Meta and others, with the goal of mitigating the risks that may arise from the use of artificial intelligence and ensuring US competitiveness in this field on a global scale.
Do not forget to follow it Xiaomi-miui.gr on Google News to be informed immediately about all our new articles! You can also if you use RSS reader, add our page to your list by simply following this link >> https://news.xiaomi-miui.gr/feed/gn
