European countries have reached an interim agreement on important EU rules on the use of artificial intelligence in measures such as facial recognition and how to regulate artificial intelligence systems such as GPT chat.
Publish dateMonday 11 December 2023 - 14:01
Story Code : 281985
Afghan Voice Agency (AVA): According to Reuters, with this landmark agreement, the European Union will become the first major world power to enact rules governing artificial intelligence.
The agreement was reached on Friday, December 8, between EU countries and members of the European Parliament after nearly 15 hours of negotiations. The two sides of this agreement are going to finalize the details in the coming days.
The agreement requires underlying models such as GPT Chat and General Purpose Artificial Intelligence (GPAI) systems to comply with transparency obligations before they can be brought to market. These applications are required to provide technical documentation, compliance with EU copyright law and detailed summaries about the content used.
Models with a higher probability of systemic risk must also report to the EU model assessments, assessment of systemic risk mitigation, cyber security assurance and energy efficiency.
It is also stipulated that governments can only use this tool for essential biometric surveillance and facial recognition in public spaces, prevention of current or foreseeable threats such as terrorist attacks, and searches for people suspected of serious crimes.
This agreement prohibits the use of artificial intelligence technology for the purpose of social scoring and use in classification systems by inferring political, religious, philosophical, sexual and racial opinions.
At the same time, it has been emphasized that there is a right to complain about the misuse of these tools. Fines for violations range from 7.5 million euros or 1.5% of global turnover to 35 million euros or 7% of global turnover.
The agreement was reached on Friday, December 8, between EU countries and members of the European Parliament after nearly 15 hours of negotiations. The two sides of this agreement are going to finalize the details in the coming days.
The agreement requires underlying models such as GPT Chat and General Purpose Artificial Intelligence (GPAI) systems to comply with transparency obligations before they can be brought to market. These applications are required to provide technical documentation, compliance with EU copyright law and detailed summaries about the content used.
Models with a higher probability of systemic risk must also report to the EU model assessments, assessment of systemic risk mitigation, cyber security assurance and energy efficiency.
It is also stipulated that governments can only use this tool for essential biometric surveillance and facial recognition in public spaces, prevention of current or foreseeable threats such as terrorist attacks, and searches for people suspected of serious crimes.
This agreement prohibits the use of artificial intelligence technology for the purpose of social scoring and use in classification systems by inferring political, religious, philosophical, sexual and racial opinions.
At the same time, it has been emphasized that there is a right to complain about the misuse of these tools. Fines for violations range from 7.5 million euros or 1.5% of global turnover to 35 million euros or 7% of global turnover.
avapress.com/vdcc4sqip2bq4o8.-ya2.html
Tags
Top hits
