Unsettling prevalence of AI-driven naked image bots and the battle against manipulative deepfake misbehavior
===============================================================================================
In 2019, an app named DeepNude, which used generative adversarial networks (GANs) to create fake nude images of women, caused a stir due to its controversial nature. The app's removal was prompted by public outcry, but its successor, a bot on Telegram, continues to pose a significant threat.
This bot, which is accessible and easy to use, has been estimated by Sensity AI to have been used to target at least 100,000 women as of July 2020, with a significant portion suspected to be underage. The bot's ecosystem includes Telegram channels dedicated to sharing and "rating" the generated images, creating a system that incentivizes users to target more individuals and share their creations.
Deepfakes introduce the possibility of creating entirely fabricated yet highly realistic content, making it more challenging for victims to seek justice or recourse. The use of deepfakes for malicious purposes, such as creating non-consensual intimate imagery, adds a new layer of complexity to the issue of revenge porn.
Addressing the proliferation of AI-powered "undressing" bots on platforms like Telegram involves a combination of technical, legal, and community-based strategies focused on preventing non-consensual image manipulation and deepfake abuse.
- Platform Moderation and Restrictions: Platforms like Telegram need to intensify content moderation to detect and block AI "undressing" bots or nudify tools. Due to their popularity (with millions of users monthly), tighter platform policies and automated detection systems are crucial to curb access.
- Legal and Regulatory Measures: Many governments are introducing or strengthening laws against non-consensual synthetic intimate imagery (deepfakes). For example, Australia’s eSafety Commissioner actively regulates online harms including deepfake abuse, providing a model for combining criminal and civil law enforcement alongside dedicated regulatory agencies. Additionally, certain regions are enforcing stricter content controls on adult and mainstream platforms to ensure accountability for hosting such content.
- Industry Actions: Major tech companies and platforms have started banning advertisements promoting deepfake pornography or tutorials on creating it. Google banned ads on such content and restricted use of its platforms in training deepfake models. Adult content websites like Pornhub pledged to ban AI-generated non-consensual porn and are subject to regulations that require cracking down on harmful content or face fines and suspensions. These industry-level safeguards help choke off demand and availability.
- User Education and Awareness: Because “undress” AI bots violate consent by producing fabricated explicit images that can be used for bullying or blackmail, public education on privacy risks, safe social media practices, and how to report abuses is critical.
- Technical Countermeasures and Research: Developing and deploying AI tools capable of detecting deepfake and manipulated images can help platforms identify and remove abusive content promptly. Research into watermarking, image provenance, and forensic analysis supports this effort.
In summary, combating the spread of AI-powered undressing bots and deepfake abuse requires coordinated efforts involving platform governance, legal frameworks, industry responsibility, user education, and advancing technical detection methods to protect individuals’ privacy and dignity online.
- The use of blockchain technology can aid in the detection and prevention of the distribution of deepfake images, as it can provide an immutable record and help trace the origin of the content.
- As AI continues to advance in the realm of art and creativity, it is important for the general news media to cover the impact of AI applications like deepfakes and their implications on cybersecurity and crime and justice.
- With concerns about privacy and consent growing, technology companies and artificial-intelligence developers should prioritize the implementation of ethical principles and safeguards in their AI systems to prevent misuse and abuse.
- Collaboration between governments, tech companies, and civil society organizations is essential in addressing the broader societal challenges presented by the rise of AI-powered technology in areas such as crime and justice.
