EU Faces High-Stakes Battle Over AI Code as Nobel Laureates Warn Against Industry Pressure

The European Commission will stage a last-ditch workshop this week to convince AI companies to sign its Code of Practice before the August 2 deadline, as Nobel laureates, academics and civil society organizations warn against bowing to industry pressure.

Major tech firms continue resisting the voluntary rules amid transatlantic tensions. The July 2-3 workshops will present the final code, including copyright provisions that haven't been circulated during the nine-month drafting process.

The resistance comes as a coalition including Nobel laureates Daron Acemoglu and Geoffrey Hinton, UC Berkeley's Stuart Russell, a CAIDP board member, and 40+ academics and civil society organizations sent a June 25 letter to Commission President Ursula von der Leyen, warning that "major GPAI model providers have drastically scaled back transparency and the rigour of safety-testing around model releases."

The Center for AI and Digital Policy, a plenary participant, warned the March draft "moves further away from the spirit of the EU AI Act" by making fundamental rights assessments optional rather than mandatory.

The coalition argued that weakening rules "shifts the burden of AI Act compliance to startups and SMEs." They called for mandatory third-party testing and scaling the AI Safety unit to 100 staff.