Spain Approves AI Bill With Heavy Fines for Unlabeled Deepfakes
Spain has become one of the first member countries of the European Union to implement its AI Act, approving a new AI bill that outlines significant penalties for unlabeled AI-generated content, particularly deepfakes.
The bill, which addresses various malicious AI practices, specifies that the failure to properly label any AI-generated or manipulated content that depicts real or nonexistent individuals is considered a serious infringement. According to Spain’s Ministry of Digital Transformation, such content must be identified as AI-generated “in a clear and distinguishable manner no later than the time of the first interaction or exposure,” aligning with the regulations set forth in the EU AI Act.
EU AI Act Overview
Initially introduced in 2021, the EU AI Act underwent extensive deliberation before being passed in March 2024. It came into force in member states in August 2024 and will be fully applicable by August 2, 2026, allowing entities time to comply with its requirements.
The Act establishes risk-based rules that AI developers and deployers must follow regarding specific AI applications and prohibits the commercialization of certain AI uses. One notable provision prohibits the use of biometric data for training algorithms utilized in criminal profiling by law enforcement, marking a significant step towards minimizing bias in AI technologies. However, exemptions for national security and border control agencies have raised concerns about balancing security and individual rights.
Key Provisions of the Spanish AI Bill
The Spanish AI bill stipulates fines of up to 35 million euros or 7% of a company’s global annual turnover for improper labeling of AI content. Additionally, fines may be imposed for the failure to implement human supervision of AI systems that use biometrics for various industrial applications, with penalties ranging from €500,000 to €7.5 million or between 1% and 2% of global turnover.
AI and Deepfake Regulation in India
In contrast, AI regulation in India remains inconsistent, with the government oscillating between establishing a comprehensive AI framework and maintaining a light-touch regulatory approach. Recently, the Indian government proposed the creation of an AI Governance Board to oversee and authorize AI applications, but deepfake regulation has not been adequately addressed.
During the last general and state elections, numerous political figures and parties misused deepfakes for misinformation campaigns. In response, the Indian government reiterated the existing IT rules, with some officials suggesting that legislation regarding deepfakes may emerge if deemed necessary.
This development in Spain highlights the increasing global scrutiny and regulatory efforts aimed at managing the implications of AI technology, particularly concerning deepfake content and its potential misuse in various contexts.
