Neuramancer AI Solutions GmbH is developing a proprietary deepfake detection technology that analyzes statistical artifacts in image and video noise—regardless of the visual content. This approach is combined with so-called probabilistic AI methods to ensure a high degree of interpretability in the analysis results. Co-founder Anatol Maier originally developed the methodology as part of his research at Friedrich-Alexander University Erlangen-Nuremberg (Chair of IT Security) and is refining the algorithms as part of a SPRIND Challenge, an innovation competition organized by the Federal Agency for Leap Innovations. It enables particularly robust identification of AI and image manipulations.
