Europe turns the dial on AI

The European Commission has set a new cadence for AI governance and innovation, pairing practical compliance tools with a strategic investment in science. 

The growing need to distinguish synthetic from authentic content is at the heart of the transparency agenda. A newly announced Code of Practice is being developed by independent experts appointed by the European AI Office. It will guide providers of generative and interactive AI in meeting the AI Act’s transparency obligations. These obligations include machine readable labelling of AI generated audio, images, video, and text. It also aims to assist deployers in making clear, and context appropriate disclosures, particularly when informing the public on matters of public interest. The code is intended to serve as a practical bridge between regulatory principle and technical implementation, with transparency duties set to take effect in August 2026.

To reinforce accountability, the Commission has also published an incident reporting template for general purpose AI models that pose systemic risk. The template is designed to standardise reporting under Article 55 of the AI Act, helping providers demonstrate compliance with commitments under the GPAI Code of Practice. The Commission has moved in tandem on high risk systems, where draft guidance and a reporting template for serious AI incidents are out for stakeholder input, signalling an effort to establish consistency across the Act’s risk tiers. 

Looking beyond compliance, the launch of the Resource for Artificial Intelligence Science in Europe (RAISE) marks an intention to align Europe’s scientific leadership with its technological sovereignty. RAISE will operate as a virtual institute which coordinates AI enabled research by providing access to AI “Gigafactories,” targeted data initiatives, substantial investments in talent, and scaled research funding under the Horizon Europe programme. Backed by €107 million of funding, the pilot phase will be complemented by plans to double annual AI investments and further develop the initiative under the next EU budget cycle. The European Research Area will provide a platform for coordination among Member States. 

These three new measures all reflect the Commission’s recent shift away from addressing the commercial advent of AI through increased regulation alone.  They include practical tools to ensure nearer term compliance with the AI Act’s transparency and reporting expectations, and provide a long term strategy to embed AI into the core of scientific discovery in Europe.