The European Union recently released the first draft of the Code of Conduct for General Artificial Intelligence (GPAI) models, which aims to standardize the development of artificial intelligence and provide companies with clear compliance guidelines to avoid huge fines. The Code covers key areas such as transparency, copyright, risk assessment and risk mitigation, and sets tough penalties for non-compliance. The editor of Downcodes will explain this important document in detail for you.
GPAI is a trained artificial intelligence with a total computing power of over 10²⁵ FLOPs. Companies expected to be subject to EU guidelines include OpenAI, Google, Meta, Anthropic and Mistral.
The document addresses several core areas for GPAI framers: transparency, copyright compliance, risk assessment, and technical/governance risk mitigation. The regulation emphasizes transparency in AI development and requires AI companies to provide information about web crawlers used to train models.
The risk assessment section aims to prevent cybercrime, widespread discrimination and loss of control over artificial intelligence. AI manufacturers should adopt a Security Framework (SSF) to break down their risk management policies and mitigate them accordingly based on systemic risks.
Companies that violate the Artificial Intelligence Act will face severe penalties and may be fined up to €35 million (currently $36.8 million) or up to 7% of annual global profits, whichever is higher.
The EU invites stakeholders to submit feedback via the dedicated Futurium platform by November 28 to help improve the next draft. The rules are expected to be finalized by May 1, 2025.
The EU's move aims to balance artificial intelligence innovation and risk management and provide a strong guarantee for the healthy development of artificial intelligence. The final draft of this code deserves close attention from the industry, and its impact will affect the global artificial intelligence industry. The editor of Downcodes will continue to follow up and report on relevant progress.