In a significant move, the European Union (EU) has adopted historic regulations on artificial intelligence (AI), paving the way for the most ambitious set of standards to date to control the use of this revolutionary technology. After nearly 38 hours […]
In a significant move, the European Union (EU) has adopted historic regulations on artificial intelligence (AI), paving the way for the most ambitious set of standards to date to control the use of this revolutionary technology.
After nearly 38 hours of negotiations between lawmakers and political leaders, an agreement to support the “AI Act” was reached last Friday.
The “AI Act” is a groundbreaking global initiative, creating a unique legal framework for the development of AI that can be relied upon for both the security and fundamental rights of individuals and businesses. It is a commitment outlined in our political guidelines and one that we have fulfilled. Today’s political agreement is indeed welcome.
Efforts to adopt the “AI Act,” which was first proposed by the EU executive in 2021, gained momentum after the emergence of OpenAI’s ChatGPT last year, bringing the potential of artificial intelligence closer to the broader public.
Regarded as a global standard, the law aims to enable governments to harness the potential benefits of AI while safeguarding against risks ranging from misinformation and job losses to copyright infringements.
The regulation, which was delayed due to disagreements surrounding the regulation of language models that collect data from the internet and the use of AI by police and intelligence agencies, will now be forwarded to member states and the EU Parliament for approval.
According to the law, technology companies operating in the EU will be obligated to disclose the data used to train their AI systems and conduct product testing, especially for high-risk applications such as autonomous vehicles and healthcare.
The law prohibits uncontrolled collection of images from the internet or surveillance cameras to create databases for facial recognition. However, exceptions are made for the real-time facial recognition used by judicial authorities for counterterrorism and serious crime investigations.
Companies that violate the law may face financial penalties of up to seven percent of their global revenues, depending on the offense and the size of the company.
The EU law is considered the most comprehensive effort to date in regulating artificial intelligence amidst a growing mosaic of guidelines and regulations worldwide.
In the United States, President Joe Biden issued an executive order in October focusing on the impact of AI on national security and discrimination, while China has implemented regulations requiring AI to reflect “socialist core values.”
Other countries, such as the United Kingdom and Japan, have largely maintained an approach that does not involve a large number of regulations.