The US government has urged the European Union to revise its proposed code of practice for AI.

This voluntary code, still being developed, seeks to align tech companies with the upcoming EU AI Act.

Non-compliance could result in penalties, including fines of up to 7% of annual sales for companies and 3% for advanced AI developers.

Critics contend that the guidelines exceed the law’s scope and impose additional burdens on developers.

The US called the proposed code too strict and offered tech experts for further talks, but has not issued formal public comments.

The code is being drafted collaboratively by tech firms, copyright stakeholders, and civil society groups under the European Commission’s supervision.

It is slated for finalization and adoption next month, incorporating input from EU institutions and member states.

🔗 Source: Bloomberg

🧠 Food for thought

1️⃣ EU’s cautious tech regulation is rooted in historical data abuses

The EU’s stringent approach to AI regulation follows a consistent pattern of European precautionary tech governance that began decades earlier.

The EU’s General Data Protection Regulation (GDPR), implemented in 2018, was influenced by historical concerns about data misuse, particularly stemming from the Nazi regime’s use of data collection to identify and persecute minorities1.

As law professor Anu Bradford noted, the EU’s regulatory approach reflects deep historical lessons about how unregulated data collection facilitated discrimination and persecution1.

This historical context helps explain why European regulators consistently prioritize preventing potential harms in their approach to AI governance, applying the same precautionary principles that shaped GDPR.

The AI Code of Practice, like GDPR before it, demonstrates how Europe’s regulatory approach is shaped by historical experiences that the United States hasn’t shared, leading to fundamentally different views on appropriate oversight.

2️⃣ US-EU tech regulation tensions reflect deeper philosophical divides