The Spanish government recently passed a landmark new bill aimed at strengthening the regulation of artificial intelligence (AI) generated content, especially for "deep falsification" content that is not correctly marked. This move is seen as an important step in combating false information and protecting the public’s interests.
Minister of Digital Transformation Oscar Lopez elaborated on the background and purpose of the bill in a press conference. He said the formulation of the bill refers to the European Union's Artificial Intelligence Act and particularly emphasizes strict transparency requirements for high-risk artificial intelligence systems. Lopez pointed out that although artificial intelligence technology has brought many conveniences to social development, it may also be abused and become a tool for spreading misinformation and attacking democracy.
Image source notes: The image is generated by AI, and the image authorized service provider Midjourney
Lopez further explained that the "deep forgery" technology is extremely misleading by editing or generating videos, photos and audio, disguising itself as real content. Anyone can be a victim of this technology, so the government must take strong measures to address this challenge. The new bill clearly states that failure to properly label content generated by AI will be considered "seriously illegal" and may face fines of up to €35 million (approximately US$38.2 million) or the amount calculated based on 7% of the company's global annual turnover.
In addition, the bill also prohibits potential manipulation, such as using subconscious technology to affect vulnerable groups. Lopez gives an example that some chatbots may induce gambling tendencies to participate in gambling, while some toys may guide children to engage in dangerous activities. These behaviors not only violate morality, but may also have far-reaching impacts on society.
Another important provision of the bill is to prohibit institutions from using artificial intelligence to classify or rated individuals through biometric data, which will affect their likelihood of obtaining social benefits or assessing the risk of crime. However, based on national security considerations, real-time biometric monitoring is still allowed in public places. This provision seeks a balance between protecting personal privacy and maintaining public safety.
The implementation of the new bill will be subject to the newly established AI regulator AESIA, but the relevant regulators will be responsible for the management in specific situations involving data privacy, crime, elections, credit ratings, insurance and capital markets. The bill still needs approval from the House of Commons before it can officially take effect. Once passed, Spain will become one of the first EU countries to implement this series of regulations.
Overall, this new Spanish bill not only provides a legal framework for the normative use of artificial intelligence technology, but also provides important reference for other countries and regions in responding to similar challenges. Through strict regulatory and transparency requirements, the Spanish government hopes to effectively prevent its potential risks and protect the public's interests while promoting technological innovation.