GPAI Enforcement Powers Now Active — AI Office Begins Compliance Dialogues With Model Providers

The European Commission's AI Office activated its enforcement powers over General Purpose AI (GPAI) model providers on August 2, 2026, following a one-year period in which providers of the most advanced models were expected to voluntarily comply. The AI Office has been conducting technical compliance dialogues since August 2025, and the August 2 activation converts those dialogues from advisory to enforceable: providers who have not notified the Commission of their most advanced models, assessed and mitigated systemic risks, or implemented the GPAI Code of Practice obligations are now subject to investigation and penalty. For law firms that have built workflows on top of GPAI-based platforms, the enforcement activation creates an indirect compliance obligation: their vendor agreements should now address GPAI compliance status, and procurement teams should confirm that the platforms they rely on have met their notification and systemic risk assessment obligations. Firms that have not yet reviewed their AI vendor contracts against GPAI requirements are carrying an unquantified supplier risk.