Madrid, March 11 (Reuters) - On Tuesday, Spain's government approved a bill that enforces hefty fines on companies utilizing artificial intelligence (AI)-generated content without proper identification to combat the spread of "deepfakes."
The bill, aligning with the European Union's guidelines, imposes stringent transparency requirements on high-risk AI systems, as stated by Digital Transformation Minister Oscar Lopez.
Lopez emphasized the dual potential of AI, stating, "AI is a very powerful tool that can be used to improve our lives or to spread misinformation and attack democracy."
Spain is among the first EU nations to adopt the bloc's regulations, viewed as more all-encompassing than the United States' system, which heavily relies on voluntary compliance and a fragmented framework.
In case of non-compliance with the correct labeling of AI-generated content, the Spanish bill categorizes it as a "serious offense," with fines reaching up to 35 million euros ($38.2 million) or 7% of their global annual turnover, pending approval by the lower house.
The legislation also prohibits other practices, like employing subliminal techniques to manipulate vulnerable groups, such as chatbots encouraging addictive behavior or toys promoting unsafe challenges, as mentioned by Lopez.
Furthermore, the bill prohibits organizations from using AI to profile individuals based on biometric data or personal characteristics to assess their eligibility for benefits or their likelihood of criminal behavior.
Nonetheless, real-time biometric surveillance in public spaces for national security purposes will still be permissible.
The newly-established AI supervisory agency AESIA will oversee the enforcement of the new regulations, excluding specific cases involving data privacy, crime, elections, credit ratings, insurance, and capital market systems, which will fall under the jurisdiction of their respective watchdogs.