DuckDuckGoose is a deepfake detection platform that primarily serves identity verification (IDV) and anti-fraud platforms, used to detect deepfake content in real-time, global scenarios such as user registration, compliance review, and trust maintenance.
Purpose
The platform is built around identity and anti-fraud workflows, detecting synthetic identities and AI-generated identity documents to prevent them from bypassing KYC and liveness detection. Its detection covers forgery forms such as selfies, documents, and voice cloning, and can be embedded into existing KYC, document, or liveness detection processes.
Main Capabilities
- Detects manipulated selfies, forged documents, and liveness spoofing.
- Provides plug-and-play SDKs and sub-second API responses that can be integrated into existing KYC, document, or liveness processes without rearchitecting and without adding user friction.
- Each detection returns explainable decisions, visual traces, and machine-readable outputs for fraud, legal, and compliance teams to act on and defend in reviews.
- Products include: Phocus (a web platform that allows review of images, videos, and audio without code, providing an exportable audit trail); Waver (real-time, language-agnostic deepfake voice detection, supporting 16+ languages, with API integration); DeepDetector (instant on-premises deepfake detection for images and videos).
- Continuously retrained against 40+ image, video, and voice generation models, with detection added within weeks after new generators are released.
- Provides detection performance data based on public benchmarks such as DFDC and FaceForensics++, as well as false positive rate metrics under real traffic.
How to Use
Users can directly inspect images, videos, and audio through the Phocus web platform without writing code; they can also integrate DeepDetector and Waver via SDKs or APIs into existing identity verification and fraud workflows. An official demo can be booked to see detection running in real time on your own KYC data.
Applicable Scenarios
Covers industries such as traditional banking, neobanks/fintech, identity verification, forensics, cryptocurrency, government and legal, insurance, and media and news. Specific use cases include deepfake biometric checks, identity document checks, morphing attack detection, and media forensics. Typical processes include KYC onboarding, insurance claims, and border control.
Limitations
The materials do not provide information such as pricing, specific deployment costs, free trials,supported languages (aside from Waver's 16+ languages), system requirements, or service-level agreements. Data such as detection accuracy and false positive rates comes from official statements, and actual results may vary depending on the environment and data.
