AI Compliance: what changed across the EU in June 2026 (3-min read)

The AI regulatory ecosystem notoriously changes fast and unforgivably: what happened in June 2026 confirmed that once more. Here’s a 3-min summary.

EU AI Act Appointments

On 1 June 2026, the EU Commission appointed the 60-member Scientific Panel and a new Advisory Forum to support the EU AI Act enforcement. The bodies will provide independent expertise and advisory to the AI Office and national authorities as to how to best apply the rules. The Scientific Panel and the Advisory Forum will focus on Frontier AI and general-purpose AI systems (GPAI).

Code of Practice

On 10 June 2026, the EU Commission published the long-awaited Code of Practice on Transparency of AI-Generated Content, a voluntary framework providing practical insights into how to implement the EU AI Act’s rules of transparency for both providers and deployers (Art. 50). In addition, the EU Commission further provided a list of icons to label AI-generated content. Read full code.

Assessing the implications of Fable’s block

On 14 June 2026, following Trump’s block of Anthropic’s Fable to non-US citizens, the EU Commission stated that it was assessing the practical implications of the US president’s executive order.

Digital Omnibus officially approved

On 16 June 2026, as expected, the EU Parliament passed the amendments to the EU AI Act, including delaying key provisions of the Regulation for high-risk systems, ultimately approving the Digital Omnibus. As such, as of 2 August 2026, Art. 50 on AI transparency will become applicable:

• Human interactions. Providers must design AI systems in a way that makes it clear to users that they are interacting with AI. The disclosure must be perceivable within the interaction layer.
• Deepfakes. Companies using AI to generate deepfakes must flag the content as synthetic via clear and distinguishable labels. Labels must be displayed at the user’s first interaction with the content in question.
• Emotions and biometrics. Providers of AI system analysing human emotions (sentiment analysis) or performing biometric categorisation (task-allocation systems) must explicitly and timely inform concerned users.*

EU retailers’ protests

On 19 June 2026, Eurocommerce, the European retail association including behemoths such as Amazon, H&M, Inditex, ‌and Ikea, asked the EU tech chief to not include AI-generated advertisements from requiring public disclosure of AI use, as mandated by the EU AI Act. The rationale being that AI-generated ads used by retail companies are not intended to mislead users, and any AI disclosure would risk diminishing the value of both the products advertised and the disclosure itself to consumer.

How to know more

Obligations will inevitable continue evolving along with the AI technology. More obligations will emerge, some might be delayed. Take the Digital Omnibus as the chief example.

A good way to keep yourself up to date with any AI governance, risk and compliance updates is to subscribe to Quantidal’s free monthly newsletter:

SUBSCRIBE HERE

*Many thanks to James Clark for the useful summary.