EU treats GPAI Code signatories as acting in good faith — Meta's refusal draws enhanced scrutiny
As enforcement begins, the EU's AI Office says the 26 providers that signed the GPAI Code of Practice — Microsoft, Google, Amazon, OpenAI, Anthropic among them — will have their commitments weighed favorably when fines are calculated. Meta, which refused to sign, faces enhanced scrutiny. The voluntary code has become the practical dividing line of the enforcement era.
The code turned into a compliance safe-harbour. By signing, 26 providers earn treatment as good-faith actors through the transition, softening how the AI Office weighs any breach; refusing, as Meta did, forgoes that cushion and invites closer examination. A voluntary instrument has quietly become the thing that determines a provider's posture on day one of enforcement.
Meta's holdout is a strategic bet with a real cost. Declining to sign preserves flexibility and avoids conceding to terms Meta considers overreaching, but it trades that for enhanced scrutiny precisely when the AI Office is deciding who to examine first. In a regime where discretion shapes outcomes, being outside the good-faith group is an expensive place to stand.
The lesson for providers is that soft law now has hard consequences. The GPAI Code carries no statutory force of its own, yet signing or not signing materially changes exposure under the binding Act — which is how a voluntary code becomes, in effect, mandatory for anyone who wants the benefit of the doubt from a regulator that now has teeth.
ComplianceHub.Wiki — EU AI Act GPAI enforcement goes live August 2, 2026: a readiness guide → · TECHi — EU AI Act GPAI enforcement begins August 2 — who is exposed? →