bbieron@platformeconomyinsights.com

US Government Challenges AI Act Code of Practice on General Purpose AI

May 5, 2025

Report from Bloomberg

In Brief – The US Government is calling on the European Commission to delay the release of the AI Code of Practice for “General Purpose” AI that is scheduled to be finalized in May to help provide guidance to developers of the largest and most capable AI models for complying with the EU’s AI Act. The code, which is being developed by the EU’s new AI regulators and multistakeholder “expert groups” that include representatives of tech firms, copyright industries, academia, and civil society groups, aims to provide AI developers with recommendations to meet the law’s standards for transparency, risk-mitigation and copyright rules that go into effect in August. Although the code is voluntary, not following the recommendations could mean greater regulatory scrutiny for an AI developer. Critics argue the code is setting guidelines that go beyond the bounds of the law and will undermine AI development by imposing onerous regulatory burdens. The US Administration’s letter to the Commission reportedly detailed concerns with both the code of practice and the AI Act at large, and called for the entire AI Act implementation process to be put on hold to address the issues.

Context – When the EU was crafting its landmark AI Act, the Parliament changed course in response to the emergence of Chat-GPT and chose to regulate large “foundation models” such as the biggest chatbots. This shift was divisive and EU-based AI innovators pushed for lesser mandates. The final version planted the flag globally for AI regulation, but that was not without risk. The USUK, and Japan have each shifted their AI policies in a deregulatory direction since last year, and there are top European leaders calling for the EU to “resynchronize with the rest of the world” to better incentivize investment and innovation in Europe. That intra-European debate is playing out over the Code of Practice, with its third and likely final draft criticized from across the spectrum of stakeholder groups, including MEPs who led the legislative effort arguing it weakened standards beyond what the bill’s drafters ever intended.

View By Monthly
Latest Blog
The Trump Administration’s Shifting View on AI Model Security Tests

Report from the New York Times In Brief – The Trump Administration is weighing a shift in policy to allow federal agencies to review advanced AI models, with officials discussing an executive order to establish an AI working group composed of government leaders and...

Big AI Industry Companies Reach Agreements with the Pentagon

Report from the Wall Street Journal In Brief – The US Defense Department has finalized agreements with eight major US technology companies to deploy their artificial intelligence tools in classified environments. The companies are OpenAI, Google, SpaceX (via xAI),...

Platform Economy Insights produces a short email four times a week that reviews two top stories with concise analysis. It is the best way to keep on top of the news you should know. Sign up for this free email here.

* indicates required