Shorouk Express
Following AI-generated political attack ads against the government in recent weeks, Spain is now moving forward with a draft bill to regulate the use of AI that require clear labelling and massive fines for those who don’t clarify that images are fake.
The Spanish government this week moved forward with draft legislation to regulate the use of artificial intelligence (AI) and adapt EU legislation into Spanish law.
The proposals seek to ensure the ethical and beneficial use of AI technology, and also clamp down on the use of deepfakes and create a new sanctioning regime.
Using AI techniques to manipulate decisions without consent, for example the identification of profiles with addiction problems, will be banned. The draft also seeks to regulate the use of AI in high-risk systems such as transport, healthcare, critical infrastructure, education, employment, migration, justice or democratic processes.
READ ALSO: Spain’s clampdown on child YouTubers and parent influencers
Óscar López, Spain’s Minister of Digital Transformation and Public Service, explained in a press conference that AI is a double-edged sword: “AI is a very powerful tool, which can be used to improve our lives or to spread hoaxes and attack democracy.”
European regulations, López added, “defend our digital rights by identifying prohibited practices’.
Advertisement
In particular, if the draft becomes law, it will be considered a serious offence not to clearly label all texts, videos or audios generated with AI. “We are all susceptible to being the target of an attack of this kind. They are called deepfakes and they are prohibited,” López warned.
This comes just a week after Spain’s opposition Partido Popular (PP) posted deepfake videos generated by AI in order to launch political attacks on the government.
READ MORE: Spanish ‘Corruption Island’ attack ad angers Dominican Republic
AI images of PM Pedro Sánchez, former Transport Minister José Luis Ábalos y prosecutor Álvaro García Ortiz, all accused of corruption in the spoof “Temptation Island” video published by the opposition PP party.
Notably, the PP posted an one AI-made attack ad called La isla de las Corrupciones (The Island of Corruption) that saw Prime Minister Pedro Sánchez, his wife, Begoña Gómez, and other figures close to Sánchez caught up in alleged corruption scandals, including former minister José Luis Ábalos and his advisor Koldo García, against a backdrop of footage of the Dominican Republic.
READ ALSO: What’s the latest on Spain’s Koldo corruption case and is the PM involved?
Advertisement
The video was withdrawn following protests from the Dominican government, which considered it to be “an attack” on the country. The PP then followed it up with another AI-made video on 8M, International Women’s Day. Neither video was identified as being generated by AI.
López explained that according to the new draft rules, “deepfakes must be identified ‘in a clear and distinguishable manner at the latest on the occasion of the first interaction or exposure.’”
However, so far the government is yet to specify how exactly these sorts of videos will be identified and marked.
A number of different bodies will oversee the implementation of AI regulations in Spain, including the Spanish Data Protection Agency and the recently founded Spanish Agency for the Supervision of Artificial Intelligence (AESIA) based in A Coruña.
Companies that fail to comply with regulations could face fines of up to €35 million and/or between 5 percent and 7 percent of global turnover.
López also noted that, according to various independent estimates, Spain’s GDP could grow by 8 percent in the next 10 years with AI alone, which amounts to €100 billion.
READ ALSO: The ugliest cities in Spain according to AI
