I’ve been deepfaked: What do I do?

This article addresses the growing problem of deepfakes and the increasing difficulty of removing them from online platforms. It outlines steps individuals can take if they've been deepfaked, focusing on preserving evidence and utilizing specialized resources like StopNCII.org for non-consensual intimate imagery. The article details reporting procedures on major platforms (Google, Facebook, Instagram, TikTok, YouTube, X) and emphasizes the importance of providing specific details about why the deepfake is harmful, rather than simply stating it's AI-generated. It also highlights the need to contact website publishers and local data protection regulators when platforms fail to remove content.

Deepfakes, created using generative AI tools, are becoming increasingly common and sophisticated, leading to a surge in instances of harassment, fraud, and impersonation. The article acknowledges the legal complexities surrounding deepfakes, noting that while some actions are criminalized (e.g., non-consensual intimate imagery), the definition of what constitutes a violation can vary significantly depending on jurisdiction.

**What happened** Deepfakes are now a prevalent issue, driven by advancements in AI technology. Cybercriminals and bullies are exploiting these tools for malicious purposes, including extortion and harassment. The article stresses the need for a multi-pronged approach, combining individual action with platform accountability.

**Technical details**

• **Generative AI:** The technology behind deepfakes is rapidly evolving, making detection and removal increasingly challenging.
• **Non-Consensual Intimate Imagery (NCII):** Legal frameworks like the TAKE IT DOWN Act and the Data Act are designed to address NCII, requiring platforms to respond swiftly to reports.
• **StopNCII.org:** This resource utilizes hashing technology to identify and remove deepfake images of victims, working with platforms like Meta, TikTok, Reddit and X.
• **TakeItDown.NCMEC.org:** A similar system for images/video of those under-18.

**Impact** The article highlights the significant real-world impact of deepfakes, including emotional distress, reputational damage, and potential legal consequences for victims. The spread of misinformation and the erosion of trust are also major concerns.

**What to do**

• **Preserve Evidence:** Take screenshots, save URLs, and document the date and time of discovery.
• **Report to Platforms:** Follow the specific reporting procedures on each platform (Google, Facebook, Instagram, TikTok, YouTube, X). Be detailed about why the deepfake is harmful (e.g., fraud, harassment, impersonation).
• **Contact Website Publishers:** Request removal directly from the source website.
• **Escalate to Regulators:** If platforms refuse to act, contact your local data protection regulator (UK/EU) or relevant state authorities (US).
• **Hashing:** Utilize StopNCII.org to create a hash of the offending image/video.

**Why it matters** The article underscores the urgent need for a coordinated response to combat the proliferation of deepfakes. It emphasizes that simply labeling content as AI-generated is often insufficient, and that providing specific details about the harm caused is crucial for effective action. The article serves as a practical guide for individuals affected by deepfakes and a call for greater accountability from online platforms and regulatory bodies.