EU Passed AI Act: Balancing Technology and Ethics

The EU's landmark AI Act sets a new global precedent in regulating artificial intelligence, with potential implications for the entertainment sector.

Europe has taken a significant step in artificial intelligence (AI) regulation with the passing of the new AI Act by European Union (EU) policymakers. This development marks the first instance of a legislative body enacting laws specifically tailored to govern AI. The European Commission, European Council, and European Parliament collaborated to finalize this Act, which addresses both commercial and public applications of AI technology.

Thierry Breton, the European commissioner instrumental in negotiating the AI Act, highlighted Europe's pioneering role in setting global standards. EU Commission President Ursula Von der Leyen also expressed her enthusiasm, noting that the Act will provide legal certainty and foster innovation in trustworthy AI, contributing to the development of global standards to find a balance between the potential and risks of AI. Although the Act requires final approval from the European Parliament and Council, it is expected to be operational in phases, with some parts effective from next year and the majority by 2025 and 2026.

The Act has undergone substantial revisions since its first draft in 2021. The EU's approach encompasses a broad range of AI applications, extending beyond recent concerns highlighted in writers and actors strikes, which revolved around the protection of creative rights against AI replacement.

One of the key components of the Act involves restrictions on facial recognition technology, particularly by police and government entities, with allowances only under certain safety and national security conditions. Additionally, the Act imposes new transparency obligations on developers of major general-purpose AI systems, aligning with standards set by U.S. President Joe Biden's executive order. Notably, companies violating these regulations could face penalties up to 7 percent of their global sales.

The real test for the AI Act will be its enforcement. Similarly to the General Data Protection Regulation 2016 (GDPR), which faced varied implementation across EU nations, the AI Act's impact, especially in different industries including entertainment, will largely depend on the enforcement across the EU. Businesses impacted by this legislation, particularly major tech companies, are likely to challenge some of its provisions, potentially leading to further delays in its continent-wide implementation.

Barry Scannell, an AI legal expert based in Ireland, emphasized the challenges businesses may face, particularly in balancing enhanced transparency requirements with the protection of intellectual property. Meanwhile, the Computer and Communications Industry Association in Europe (CCIA) criticized the Act for potentially over-regulating AI, expressing concerns over stifling innovation and driving AI companies and talent away from Europe.

On the other side of the debate, civil rights groups have raised concerns that the legislation does not go far enough, particularly in the regulation of AI-assisted facial recognition technology by governments and police. Amnesty International's Mher Hakobyan criticized the EU's decision for not fully banning facial recognition, highlighting the potential risks to human rights, civic space, and the rule of law.

Although the EU AI Act is not specifically pertaining to the entertainment sector, as an overall and comprehensive regulation framework on the technology, it still impacts the entertainment industry, with both direct and indirect effects. The entertainment companies, in particular operations in the EU, will need to adapt to the broader legal and ethical framework established by the Act, especially in areas like data usage and AI-driven content creation. Thus again, while the EU has taken a first step, there is room for negotiations on details for AI companies, and the impact on broader industries remains dependent on the implementation of the Act and whether more detailed and other industry-focused regulations might come in the future.


Previous
Previous

Axel Springer Partners with OpenAI

Next
Next

MARZ's LipDub AI Revolutionizes Dubbing for Global Audiences