OpenAI has joined the Coalition for Content Provenance and Authenticity (C2PA) steering committee to increase transparency around generated content. The C2PA standard certifies digital content with metadata proving its origins, whether created by AI or through traditional means. OpenAI is integrating C2PA metadata into its generative AI models, starting with its DALL-E 3 model output. This move aims to combat the spread of AI-generated misinformation, especially during major elections. OpenAI is also developing new provenance methods like watermarking for audio and image detection classifiers. The company has opened applications for access to its DALL-E 3 image detection classifier through its Researcher Access Program. Additionally, OpenAI is incorporating watermarking into its Voice Engine custom voice model. The company believes that increased adoption of provenance standards will enhance digital content authenticity practices. OpenAI is also collaborating with Microsoft to launch a $2 million societal resilience fund to support AI education and understanding. The company emphasizes the importance of collective action in enabling content authenticity and promoting transparency online.
Source link
Source link: https://www.artificialintelligence-news.com/2024/05/08/openai-steps-boost-ai-generated-content-transparency/
OpenAI enhances AI content transparency; promotes trust with #transparency
