EU calls for collaboration to shape regulation of AI use | Festina Lente - Your leading source of AI news | Turtles AI

European Union Launches Consultation on Code of Conduct for AI Models.

 Key Points:
1. Purpose of the Code of Conduct: To define rules for providers of general-purpose AI models.
2. Timeline for Implementation: The Code will be implemented by April 2025.
3. Broad Participation: Consultation is open to GPAI providers, companies, civil society, academics and experts.
4. Consultation Details: Focus on transparency, risk assessment, and monitoring of the Code.

The European Union has launched a public consultation aimed at establishing a Code of Conduct for providers of general-purpose artificial intelligence (GPAI) models, such as Anthropic, Google, Microsoft, and OpenAI. This process is part of the AI Act, which was recently adopted and will be effective August 1, 2024. The AI Act aims to create a risk-based regulatory system for artificial intelligence applications.

The consultation, which will run until September 10, 2024, is aimed at a wide range of stakeholders, including GPAI providers operating in the EU, companies, civil society representatives, rights holders, and academics. The goal is to gather feedback to develop a Code of Conduct to guide AI providers in meeting their legal obligations, while ensuring that GPAI is trustworthy and safe.

The Code of Conduct will be developed in an iterative process, with a first draft expected by April 2025. This will include rules on transparency, copyright provisions, risk assessment and mitigation, especially for GPAIs with systemic risks, i.e., those that exceed certain computational capacity thresholds.

In addition, the EU AI Office will play a crucial role in overseeing the process. The office is responsible for coordinating joint investigations to identify possible noncompliance and supporting national competent authorities in developing "regulatory sandboxes," controlled environments to safely test AI solutions.

The call for consultation also provides an opportunity to influence the template for the content summary used for AI model training,ensuring that the information is sufficiently detailed to allow stakeholders to exercise their rights.

Expressions of interest to participate in the drafting of the Code must be submitted by August 25, 2024. Selected participants will have the opportunity to take part in workshops and virtual meetings to help shape the Code. These meetings will be documented to ensure transparency of the process.