Stability AI's open-source Stable Diffusion image generation model was used to create non-consensual pornographic deepfakes of celebrities and other explicit content after being released without effective content filters.
Stability AI released Stable Diffusion, an open-source AI image generator capable of producing realistic pictures from text prompts that can run on consumer hardware. The model leaked early on the 4chan discussion board where it was used to generate nude celebrity images and other pornographic content. While Stability AI included a Safety Classifier tool to detect and block offensive images, this filter can be disabled by users. Unlike other AI art systems like OpenAI's DALL-E 2 which have strict filters, Stable Diffusion's open-source nature and ability to create images of public figures enables the creation of pornographic deepfakes. The report notes that 90-95% of deepfakes are non-consensual and about 90% target women. Experts warn that the system's ease of use and minimal resource requirements could enable personalized blackmail attacks and scaled harassment campaigns. Several content platforms including OnlyFans and Patreon have policies against deepfakes but enforcement remains challenging.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to gain a personal advantage over others such as through cheating, fraud, scams, blackmail or targeted manipulation of beliefs or behavior. Examples include AI-facilitated plagiarism for research or education, impersonating a trusted or fake individual for illegitimate financial benefit, or creating humiliating or sexual imagery.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed