European Union Drafts Law To Regulate Artificial Intelligence
If enacted, the law would ban real-time facial recognition and set up transparency standards for generative AI tools such as ChatGPT.
- The European Parliament has approved the first draft of its new AI Act. The new law focuses on developing and using AI systems per international human rights laws.
- This will likely set an example for other countries looking to set their own regulations. China, for instance, is expected to create a draft of its AI law this year.
The European Union has taken steps to be one of the first to pass laws to govern the use and development of artificial intelligence, putting it ahead of other countries, including the U.S. With the process initiated in 2021, the European Parliament passed the first draft of a new AI Act that aims to ensure that AI tools operate according to international human rights.
The law is expected to be revised and set into action by the end of the year after negotiations between the Council of the EU, the Parliament, and the European Commission. The first draft covers generative AI systems such as ChatGPT, Google Bard, or Bing, primarily targeting the prevention of biased, racist, and illegal content.
The new law states that AI systems will be analyzed and classified according to the level of risk posed to users, which in turn will result in more or less regulation:
- AI systems deemed to manipulate cognitive behavior, classify people based on socioeconomic status and personal characteristics, and have real-time biometric identification will be banned.
- AI systems under the EU’s product safety legislation or those associated with law enforcement, immigration, public services, infrastructure, education, and legal assistance will be categorized as high-risk and operate under continuous supervision.
- AI systems with limited risks and generative AI tools must comply with basic transparency requirements. These include notifying users when they interact with AI, highlighting content generated by AI, creating summaries of copyrighted material used to train AI models, and measures to prevent the creation of illegal content.
See More: OpenAI Sued as ChatGPT Falsely Accuses Man of Embezzlement
Chinese Government Plans To Create Draft of AI Law Before 2024
Governments worldwide have raised concerns about standards that should determine the regulate artificial intelligence. The U.S. witnessed the announcement of three bills aimed at AI transparency last week and proposed data security and privacy rules.
Furthermore, the 2023 legislation plan of the State Council in China included a draft AI law. In standard cases, China’s national legislature reviews each draft three times before voting.
In April, the Cyberspace Administration of China (CAC) announced draft rules to regulate services similar to ChatGPT. It may make it mandatory for developers to pass security assessments before such AI tools are publicly launched. This will eliminate suggestions associated with regime subversion and violence.
These moves reflect the global urgency to create adequate infrastructure to keep the rapid development of tools like ChatGPT under control.
What other important factors should governments consider when regulating AI? Let us know your thoughts on LinkedIn, Twitter, or Facebook.
Image source: Shutterstock
LATEST NEWS STORIES
