Policy Backgrounders

Proposal for a US-EU AI Code of Conduct

June 09, 2023

The Fourth Ministerial meeting of the US-EU Trade and Technology Council concluded with officials stating their intention to develop a voluntary AI code of conduct, filling the interim gap as the EU moves towards passage of its landmark standard-setting Artificial Intelligence Act.

• A draft voluntary AI code of conduct is expected within the coming weeks and will focus on transparency and risk auditing.

• The EU’s AI Act first introduced in 2021 was approved by the Council of the EU in December 2022. The AI Act is now scheduled for a vote on June 14 by the European Parliament before negotiations would commence to progress a final form of the law for adoption by EU member states, potentially by the end of 2023 before it would go into force likely in 2026.

• The US and EU agree on general AI principles, including a risk-based approach with regulation proportionate to the level of risk posed by the type of AI system. The shared US-EU code of conduct suggests that US officials may look to develop a US framework that aligns to EU standards, which are now largely established and undergoing fine-tuning before passage.

• US and EU officials plan to release a draft voluntary code of conduct within the coming weeks. Private industry is expected to be able to submit feedback on that draft framework before being presented to G7 leaders as a joint transatlantic proposal in the fall. That voluntary code of conduct will allow companies to sign on and be focused on transparency, risk auditing, and other technical details around development of AI systems.