Spain’s government approved a bill on Tuesday imposing massive fines on companies that fail to label content generated by artificial intelligence (AI). This legislative move aims to curb the proliferation of so-called ‘deepfakes’ and enhance transparency regarding AI use.
The bill follows the guidelines set by the European Union’s landmark AI Act, which mandates stringent compliance for high-risk AI systems. Digital Transformation Minister Oscar Lopez commented during a press conference, ‘AI is a very powerful tool that can be used to improve our lives … or to spread misinformation and attack democracy.’
Notably, Spain stands out as one of the first countries in the EU to implement these comprehensive regulations, surpassing the mainly voluntary compliance model adopted by the United States. Lopez highlighted that ‘everyone is susceptible to deepfake attacks’—referring to manipulated videos, images, or audio presented as genuine content.
Once approved by the lower house, the bill categorizes the failure to adequately label AI-generated content as a ‘serious offence.’ Companies could face fines reaching up to €35 million (approximately $38.2 million) or 7% of their global annual turnover.
This regulatory effort comes in response to concerns over the societal harm posed by unregulated AI technologies, particularly since the introduction of OpenAI’s ChatGPT in late 2022, which demonstrated the capabilities of AI to engage users with human-like interactions.
In addition to labeling requirements, the legislation prohibits the use of subliminal messaging techniques aimed at manipulating vulnerable demographics, such as exploiting gambling addictions or encouraging risky behavior among children.
The bill also restricts organizations from utilizing biometric data to profile individuals for benefits or crime risk assessment, though it permits real-time biometric surveillance in public spaces for national security purposes.
The newly established AI supervisory agency, AESIA, will oversee the enforcement of these rules, apart from specific areas related to data privacy, law enforcement, election processes, and financial systems, which will remain under the jurisdiction of relevant authorities.
