As the EU considers new regulations for AI, 60+ NGOs ask the EC to ban automatic gender recognition systems, which rely on outmoded understandings of gender
Dangers posed by facial recognition like mass surveillance and mistaken identity have been widely discussed in recent years.
Context & Ripple Effects
This letter lands mid-arc in a two-year civil-society campaign against biometric AI in Europe. An [[a:943185|EU expert panel had already recommended banning AI-based social credit scoring and limiting mass surveillance]] back in 2019, and the NGOs' gender-recognition demand arrives just as the Commission drafts the rules that would become the AI Act.
The campaign then escalates rather than fades: weeks later an [[a:1159574|open letter with 170 signatories in 55 countries calls for a full ban on biometric recognition tech]], arguing the proposed rules have loopholes, and by September more than [[a:970292|300 organizations urge the EU to relabel more AI systems as high-risk and prohibit some outright]]. The eventual outcome — lawmakers in 2023 agreeing to ban facial recognition in public spaces — shows these letters were shaping the final text, not shouting into the void.
First-order effects
- The European Commission faces direct pressure while drafting the AI Act to classify automatic gender recognition as a prohibited practice rather than merely high-risk, since the NGOs argue the systems encode outmoded understandings of gender.
Second-order effects
- Vendors selling gender-classification features inside facial recognition products confront a shrinking addressable market in the EU, pushing them toward other jurisdictions or dropping the capability entirely.
- Each successive open letter raises the coordination bar for industry lobbying, since opponents of a ban must now argue against a broadening civil-society coalition rather than isolated complaints.
Third-order effects
- If the pattern holds, the EU institutionalizes rights-based red lines — categories like gender that regulators deem unsuitable for automated inference get banned outright instead of audited, setting a template other jurisdictions copy or deliberately reject.
The trend: EU AI regulation is being pulled from 'manage risky systems' toward flat prohibition on biometric inference, with coordinated NGO campaigns acting as the forcing function.