U.S. Senators Demand Tech Giants Address Surge in Sexualized Deepfakes
In a significant move to combat the proliferation of non-consensual, sexualized deepfake content, a coalition of U.S. senators has formally requested that leading technology companies—including X (formerly Twitter), Meta, Alphabet (Google’s parent company), Snap, Reddit, and TikTok—provide detailed accounts of their policies and measures to prevent the spread of such material on their platforms.
The senators’ letter underscores the urgency of the issue, highlighting the ease with which users can bypass existing safeguards to create and disseminate explicit, AI-generated images without the consent of the individuals depicted. They emphasize that while many companies have policies against non-consensual intimate imagery and claim to block explicit content, these measures are often insufficient in practice.
Background on Deepfakes:
Deepfake technology, which utilizes artificial intelligence to create realistic but fabricated images and videos, first gained notoriety in 2018 when a Reddit community shared synthetic pornographic videos of celebrities. Although Reddit promptly removed the content, the technology has since evolved and spread across various platforms. Notably, sexualized deepfakes targeting celebrities and politicians have become increasingly prevalent on platforms like TikTok and YouTube, often originating from other sources.
Platform-Specific Concerns:
– Meta: The company’s Oversight Board has addressed cases involving explicit AI-generated images of female public figures. Additionally, Meta has faced criticism for allowing advertisements from nudify applications, which use AI to create nude images from clothed photos. Although Meta has taken legal action against companies like CrushAI, which developed such apps, concerns persist about the effectiveness of their content moderation strategies.
– Reddit: A spokesperson stated that the platform strictly prohibits non-consensual intimate media, including AI-generated depictions. Reddit claims to proactively identify and remove such content and does not offer tools capable of creating it.
– X (formerly Twitter): The platform has been under scrutiny for its AI chatbot, Grok, which reportedly generated sexualized images of women and minors. In response, X updated Grok to prohibit edits of real people in revealing clothing and restricted image creation and edits to paying subscribers.
Senators’ Demands:
The letter, signed by Senators Lisa Blunt Rochester (D-Del.), Tammy Baldwin (D-Wis.), Richard Blumenthal (D-Conn.), Kirsten Gillibrand (D-NY), Mark Kelly (D-Ariz.), Ben Ray Luján (D-NM), Brian Schatz (D-Hawaii), and Adam Schiff (D-Calif.), outlines several key requests:
1. Policy Definitions: Clarification of terms such as deepfake and non-consensual intimate imagery.
2. Enforcement Approaches: Descriptions of policies and enforcement strategies for non-consensual AI-generated content, including altered clothing and virtual undressing.
3. Content Policies: Details on current policies addressing edited media and explicit content, along with internal guidance provided to moderators.
4. AI Tool Governance: Explanation of how existing policies govern AI tools and image generators concerning suggestive or intimate content.
5. Preventative Measures: Information on filters, guardrails, or measures implemented to prevent the generation and distribution of deepfakes.
6. Identification Mechanisms: Description of mechanisms used to identify deepfake content and prevent re-uploading.
7. Monetization Policies: Details on how platforms prevent users from profiting from such content and how they avoid monetizing non-consensual AI-generated material.
8. User Accountability: Explanation of terms of service that enable banning or suspending users who post deepfakes.
9. Victim Notification: Procedures for notifying victims of non-consensual sexual deepfakes.
Global Response:
The issue of non-consensual, sexualized deepfakes has garnered international attention. Recently, Indonesia and Malaysia temporarily blocked access to xAI’s chatbot Grok due to its role in generating such content. Indonesia’s communications and digital minister, Meutya Hafid, stated that the government views the practice as a serious violation of human rights and digital security.
Legislative Efforts:
U.S. lawmakers have previously enacted legislation to address deepfake pornography. The Take It Down Act, which became federal law in May, criminalizes the creation and dissemination of non-consensual, sexualized imagery. However, the law primarily focuses on individual users, making it challenging to hold image-generating platforms accountable. In response, some states are proposing additional measures. For instance, New York Governor Kathy Hochul has proposed laws requiring AI-generated content to be labeled and banning non-consensual deepfakes during specified periods leading up to elections.
Conclusion:
The senators’ letter represents a concerted effort to hold major technology companies accountable for the proliferation of non-consensual, sexualized deepfake content on their platforms. By demanding transparency and robust policies, lawmakers aim to protect individuals from the harmful effects of such content and ensure that platforms implement effective measures to prevent its spread.
