Spain’s council of ministers
adopted
proposed laws aimed at imposing hefty penalties on artificial intelligence firms failing to properly tag AI-created material with the intention of curbing “deepfakes.”
The legislation deems inaccurate AI labeling as a grave violation, carrying a possible fine ranging from €7.5 million to €35 million, or alternatively, 2% to 7% of the business’s total worldwide revenue.
The proposed law indicates that penalties for startups and medium-sized enterprises could be reduced.
“Artificial intelligence should assist us in enhancing the world we inhabit,” stated Oscar Lopez, Spain’s Minister of Digital Transformation, regarding the new legislation.
Instagram
on Tuesday.
Lopez said the law will focus AI use on medical research, preventing disasters and improving traffic in Spanish cities instead of “spreading hoaxes, lies and defamation”.
Spain’s legislation follows the guidelines set out in the EU AI Act, which introduces stringent transparency requirements for AI systems classified as high-risk, as stated by Lopez.
The bill requires approval from Spain’s Congress of Deputies before it can be implemented.
Other prohibitions imposed by the Act
The Spanish government
says
The proposed legislation prohibits other methods such as the employment of subliminal tactics, which encompass “images or sounds that cannot be perceived consciously,” aimed at influencing choices without obtaining permission.
A case in point for such techniques could be a chatbot designed to recognize individuals struggling with gambling addictions and subsequently prompting these users to access a gambling site.
The proposed legislation would additionally prohibit AI firms from categorizing individuals according to factors such as race, political affiliation, religious beliefs, or sexual orientation using either biometric information or activity on social platforms.
This information can’t be utilized to judge if someone is more prone to committing a crime compared to others, which has been pointed out as a flaw in an algorithm employed at a prison in Catalonia for assessing the risk level of inmates.
reoffending
.
AI firms utilizing biometric technology for tracking staff attendance without human oversight may face penalties ranging from €500,000 to €7.5 million, or alternatively, fines equivalent to 1% to 2% of their yearly worldwide revenue.
The legislation permits the Spanish government to temporarily remove an AI system from circulation within the nation “if it has led to a significant incident, such as causing someone’s death.”
