Analysis: 16 of the biggest so-called undress and “nudify” websites have used sign-in infrastructure from Google, Apple, Discord, X, Patreon, or Line for months
Single sign-on systems from several Big Tech companies are being incorporated into deepfake generators, WIRED found.
Context & Ripple Effects
This finding identifies a distribution-layer dependency inside a market that had already expanded rapidly: 34 undressing-app providers drew more than 24 million unique visitors in September 2023, while an earlier study counted roughly 150,000 nonconsensual porn deepfakes across 30 sites in a single month.
The relevant infrastructure is not only the image-generation tool but the account and identity layer that reduces friction for users. Later reporting that 85 nudify sites relied on Big Tech services suggests this is a recurring platform-governance exposure rather than an isolated site-level issue.
First-order effects
- The 16 identified sites gain a familiar, lower-friction account-creation path through sign-in systems operated by Google, Apple, Discord, X, Patreon, or Line.
- Those providers face immediate scrutiny over whether and how their sign-in integrations are available to services producing nonconsensual sexual deepfakes; any access review could disrupt affected sites' onboarding flows.
Second-order effects
- Nudify operators that lose or fear losing third-party sign-in options would need to rely more heavily on direct account systems, raising user-acquisition and retention friction relative to services retaining mainstream integrations.
- The report broadens the accountability question beyond model providers: identity, hosting, payments, and other intermediary services become potential pressure points, consistent with evidence that the sector's growth has been supported by services supplied by large technology companies.
Third-order effects
- If platform operators treat identity integrations as a safety-control surface, access to mainstream account infrastructure may become a more consequential boundary for abusive AI services than content moderation alone.
- The longer-run issue is whether interoperable consumer infrastructure can retain broad neutrality while responding to clearly harmful downstream uses; the answer will shape where responsibility is allocated across the AI-service stack.
The trend: AI safety enforcement is moving outward from models and content toward the platform infrastructure that makes harmful services easy to discover, join, and operate.