Tech news site Motherboard reported that a Reddit user called ‘deepfakes’ was using freely available machine learning tools to swap celebrities’ faces into pornographic videos without consent. The term ‘deepfake’ quickly became shorthand for AI-manipulated media. The videos were fake, but they were convincing enough to alarm researchers and the public.
What happened
- The user trained neural networks with open-source tools, including TensorFlow, using images gathered from search engines, stock photos, and videos.
- The process needed only a consumer-grade computer, and an easy-to-use app followed within weeks.
- Reddit, Twitter, and other platforms banned non-consensual deepfake pornography in early 2018.
- Researchers and journalists warned that the technology could be used for harassment, fraud, and disinformation.
Why it mattered
The report marked the moment AI face-swapping moved from research labs to the public, foreshadowing later deepfake fraud, impersonation, and non-consensual imagery. Deepfake technology has since become cheaper and more realistic, making it a growing concern for security teams.
Lessons for organisations
Organisations should prepare for deepfake impersonation, for example by verifying payment and sensitive requests through a separate channel, and include synthetic media in security awareness training.
Source: Motherboard (VICE)
Part of our Top stories archive of headline-making events in information security, privacy, and AI. If you would like help applying the lessons to your organisation, contact us.
