Artificial intelligence
Sections
Highlight
Sections
Highlight
Spain's central government has approved a draft bill to regulate ethical, inclusive and beneficial use of AI (artificial intelligence). The proposal for the new law, announced following the council of ministers on Tuesday 12 March, adapts to EU regulation, which already focuses on the appropriate AI usage.
"This regulation completes the great European shield of digital rights, which protects citizens in the current context," said Óscar López, minister for digital transformation and public administration. According to López, this regulation "pioneers at national and international level" and seeks to harmonise national legislation on new digital tools.
Following the announcement, the draft bill for the ethical, inclusive and beneficial use of artificial intelligence, as it has been titled, will continue to be processed through the urgency procedure, although many of the regulations are already in force.
Since last August, the European AI Act has been regulating the use of artificial intelligence where it restricts the marketing, commissioning or use of systems that use deceptive techniques or that may manipulate a person's behaviour or decision-making; that exploit vulnerabilities of age, disability or social or economic status; or the assessment of people based on their social behaviour or characteristics through scoring that may result in detrimental treatment.
The EU law also bans the use of AI systems to: predict the risk of a person committing a crime based solely on their profile or characteristics; expand facial recognition databases by non-selectively extracting facial images from the internet or CCTV footage; focus on individual categorisation of persons based on biometric data; infer emotions in work or educational settings; or facilitate 'real-time' remote biometric identification systems in public spaces for law enforcement purposes, subject to exceptions.
"Artificial intelligence is a very powerful tool, but it can have good and bad uses. With this draft bill, we identify risks and banned practices, such as subliminal techniques or biometric recognition in public spaces, as well as ensuring the labelling of deepfakes," said López.
According to the bill, the Spanish authorities responsible for supervising compliance with the law will be: the Spanish data protection agency in cases related to biometrics; the general judiciary council in cases linked to the justice sphere; and the central electoral board in cases related to the implementation of AI in electoral processes. The Spanish AI supervision agency (AESIA) will be in charge of the rest of applications.
The draft bill also outlines serious penalties, ranging from 7.5 to 35 million euros, or between 2% and 7% of the global turnover of the company in charge. The amount may be lowered for SMEs (small- and medium-sized enterprises).
Serious infringements will include failing to comply with the obligation to correctly tag any image, audio or video that is generated or manipulated with AI and that shows real or non-existent people saying or doing things they have never done or in places they have never been - the definition of deepfake.
The bill coincides with the controversy of a deepfake video spread by Spain's Partido Popular conservative party, titled La isla de las corrupciones. The content, which was deleted after protests in the Dominican Republic, shows images of the Caribbean country and members of the ruling PSOE socialist party, hinting at corruption practices. Days later, PP produced another publication, showing a succession of virtual images of various leaders or former leaders of the left, as well as president Pedro Sánchez's wife.
Publicidad
Publicidad
Publicidad
Publicidad
Esta funcionalidad es exclusiva para registrados.
Reporta un error en esta noticia
Comentar es una ventaja exclusiva para registrados
¿Ya eres registrado?
Inicia sesiónNecesitas ser suscriptor para poder votar.