Google has officially committed to the European Union's voluntary code of practice for general-purpose artificial intelligence, a framework designed to assist AI developers in aligning with the bloc's forthcoming AI Act. This decision positions Google distinctly from Meta, which earlier this month publicly declined to endorse the code, expressing apprehension that the EU's approach to AI legislation might be overly expansive and misdirected for the industry's future.
This commitment from Google emerges just days before new regulations for providers of "general-purpose AI models with systemic risk" are set to take effect on August 2nd. These regulations will impact prominent entities in the AI space, including Anthropic, Google, Meta, and OpenAI, requiring full compliance with the comprehensive AI Act within a two-year timeframe. While acknowledging improvements in the final version of the code, Google's Kent Walker, president of global affairs, conveyed lingering concerns that the AI Act and its associated code could potentially impede Europe's progress in AI development and deployment. He specifically highlighted potential risks to European competitiveness stemming from aspects related to copyright law, approval delays, and the disclosure of trade secrets.
By formally adopting the EU's code of practice, AI companies are expected to abide by a set of directives. These include providing transparent documentation for their AI tools and services, refraining from training AI systems on illicit content, and honoring requests from content creators to exclude their works from training datasets. The EU's AI Act, a risk-based regulatory framework for AI applications, establishes prohibitions on "unacceptable risk" use cases, such as manipulative cognitive behavioral practices and social scoring. Furthermore, it delineates "high-risk" applications, encompassing biometric and facial recognition technologies, as well as AI's deployment in sectors like education and employment. Developers are also mandated to register their AI systems and fulfill rigorous risk and quality management obligations.
Google's embrace of the EU's AI code of practice signifies a proactive step towards navigating the evolving global regulatory environment for artificial intelligence. This engagement, despite expressed reservations, underscores a commitment to responsible innovation and collaboration, which is crucial for fostering public trust and accelerating the beneficial integration of AI technologies into society. It sends a positive signal about the potential for industry leaders to work within regulatory frameworks, contributing to a more secure and equitable AI future.
