The AI Act and its green blind spots: Hidden environmental risks in the AI lifecycle

The European Union Artificial Intelligence (AI) Act is the first comprehensive regulatory framework for AI,
establishing a risk-based approach. Environmental impacts are not considered a separate category of risks but are
tackled indirectly through procedural mechanisms, including transparency, documentation, and governance
requirements. Many have questioned whether this approach adequately covers environmental concerns. Yet, few
examined how such claims manifest through its compliance mechanisms. The article addresses this gap through a
doctrinal and operational analysis of the Act, identifying and coding the environment-relevant provisions across
its Articles and Annexes. Particular attention is paid to implementation via conformity assessment and post-
market governance processes. The analysis highlights a tension between the Union’s constitutional commit
ment to environmental protection and the often modest, non-mandatory environment-related obligations under
the Act, scattered across procedural duties. Based on these findings, an operational reform pathway is suggested
that relies on shared standards integrating life-cycle assessment and post-market monitoring, without altering the
Act’s risk classification