Resemble AI Launches DETECT-World for Deepfake Detection on Eden AI

On August 27, 2026, Resemble AI announced the launch of its latest deepfake detection model, DETECT-World, on the Eden AI platform. This model is designed to help organizations detect deepfakes and AI-generated content through a unified API, streamlining integration for developers. DETECT-World incorporates a dual-layer detection approach, analyzing both statistical artifacts and physical consistency in media. It has been benchmarked against over 250 generative models and supports 54 languages, achieving 95.77% accuracy for images and 98.2% for videos. The integration aims to address the growing threat of deepfake attacks, which Resemble AI documented at 821 verified incidents in the first half of 2026, matching the total for all of 2025. The model is available without requiring a new vendor relationship for existing Eden AI users.

Key Points: • DETECT-World launched on Eden AI for streamlined deepfake detection. • The model achieves 95.77% accuracy for images and 98.2% for videos. • Resemble AI reported 821 verified deepfake attacks in H1 2026.