/
← Voltar ao feed
low

Guidance for deepfake victims: legal context, evidence preservation and platform takedown routes (TAKE IT DOWN Act, StopNCII)

1 fonte(s) · 1 atualização(ões) · publicado 02/09/2026 08:48 · coletado há 4 dias

This WeLiveSecurity explainer gives practical advice to people targeted by AI-generated deepfakes, which are increasingly used in fraud, extortion, harassment and non-consensual intimate imagery (NCII). It sets out the legal context: the US TAKE IT DOWN Act criminalizes NCII distribution and requires platforms to remove reported content within 48 hours. The UK Online Safety Act and related law criminalize creating NCII and impose a 48-hour removal window, and the EU AI Act will ban systems designed to create NCII. Outside NCII, legality depends on intent, for example fraud, blackmail, defamation, harassment or impersonation. The article is guidance only and describes no specific incident, threat actor or vulnerability.

Detalhes técnicos

  • Reporting paths are described for Google, Facebook, Instagram, TikTok, YouTube and X, usually by choosing the harm category (impersonation, harassment, fraud) and stating that the content is AI-generated.
  • StopNCII.org (run by the Revenge Porn Helpline, part of SWGfL) hashes the image on the victim's device. Participating platforms, including Meta, TikTok, Reddit and X, then block or remove matching content.
  • TakeItDown.NCMEC.org offers a similar service for images or video of people under 18.
  • Google's personal-information removal tool and legal removal form can de-index content from search, but this does not remove the underlying content.

Remediação

  • Preserve evidence: screenshots, URLs, account or username, date and time, and any accompanying text. Do not engage with the poster.
  • Report to the platform and, where relevant, to the police, especially if you feel physically threatened or the content involves a child.
  • If a website refuses removal, contact its legal or privacy contact. In the UK/EU, cite a GDPR Article 17 right-to-erasure request and escalate to the data protection regulator if needed. In the US, the route depends on state law.
  • Explain why the content is harmful (fraud, harassment, impersonation), not just that it is AI-generated. Contact the search engine, the site and the regulator in parallel.

Linha do tempo

  1. welivesecurity.com ↗ · pub. 02/09/2026 08:48 · coletado 19/10 18:00 primeiro registro

    Guidance for deepfake victims: legal context, evidence preservation and platform takedown routes (TAKE IT DOWN Act, StopNCII)

    ler a análise desta fonte

    This WeLiveSecurity explainer gives practical advice to people targeted by AI-generated deepfakes, which are increasingly used in fraud, extortion, harassment and non-consensual intimate imagery (NCII). It sets out the legal context: the US TAKE IT DOWN Act criminalizes NCII distribution and requires platforms to remove reported content within 48 hours. The UK Online Safety Act and related law criminalize creating NCII and impose a 48-hour removal window, and the EU AI Act will ban systems designed to create NCII. Outside NCII, legality depends on intent, for example fraud, blackmail, defamation, harassment or impersonation. The article is guidance only and describes no specific incident, threat actor or vulnerability.

    Technical Details

    • Reporting paths are described for Google, Facebook, Instagram, TikTok, YouTube and X, usually by choosing the harm category (impersonation, harassment, fraud) and stating that the content is AI-generated.
    • StopNCII.org (run by the Revenge Porn Helpline, part of SWGfL) hashes the image on the victim's device. Participating platforms, including Meta, TikTok, Reddit and X, then block or remove matching content.
    • TakeItDown.NCMEC.org offers a similar service for images or video of people under 18.
    • Google's personal-information removal tool and legal removal form can de-index content from search, but this does not remove the underlying content.

    Remediation

    • Preserve evidence: screenshots, URLs, account or username, date and time, and any accompanying text. Do not engage with the poster.
    • Report to the platform and, where relevant, to the police, especially if you feel physically threatened or the content involves a child.
    • If a website refuses removal, contact its legal or privacy contact. In the UK/EU, cite a GDPR Article 17 right-to-erasure request and escalate to the data protection regulator if needed. In the US, the route depends on state law.
    • Explain why the content is harmful (fraud, harassment, impersonation), not just that it is AI-generated. Contact the search engine, the site and the regulator in parallel.