Hive CSAM Detection API is a cloud-based content moderation solution designed to identify, classify, and report child sexual abuse material (CSAM) across images, videos, and text. The API leverages advanced machine learning models and hash matching technology, developed in partnership with Thorn and integrated with the Internet Watch Foundation (IWF), to detect both known and novel CSAM. It supports real-time and batch processing, providing risk scores and classification labels (CSAM, pornography, benign) to facilitate rapid human review and escalation. The API also includes seamless reporting workflows to the National Center for Missing & Exploited Children (NCMEC) via Hive’s Moderation Dashboard, helping platforms meet legal obligations. The solution is designed for platforms with user-generated content, enabling scalable, automated moderation and compliance with child safety regulations.

Key Features

AI Image and Video Classifier Detects new and novel CSAM using state-of-the-art machine learning models.

• Classifies content as CSAM, pornography, or benign
• Generates risk scores for prioritization

Hash Matching for Known CSAM Identifies known CSAM by matching against a database of over 57 million hashes.

• Uses cryptographic and perceptual hashing
• Detects manipulated and altered images/videos

Text-Based Exploitation Detection Classifies and scores text content for child sexual exploitation risks.

• Detects CSE in chats, comments, and messages
• Assigns risk scores across key categories

Integrated Reporting Workflow Streamlines mandatory reporting to NCMEC via Hive’s dashboard.

• Automates report submission
• Ensures compliance with legal requirements

Real-Time and Batch Processing Supports high-volume content moderation.

• Instant API responses for real-time detection
• Batch processing for large datasets

Partnerships and Trusted Data Utilizes technology and data from Thorn and IWF.

• Access to IWF’s image and video hashes
• Trained on NCMEC CyberTipline data

Benefits

Enhanced Child Safety Proactively detects and removes harmful content to protect children.

• Reduces exposure to CSAM
• Supports platform trust and safety

Regulatory Compliance Helps platforms meet legal obligations for CSAM detection and reporting.

• Integrated NCMEC reporting
• Supports global compliance standards

Operational Efficiency Automates high-volume moderation tasks.

• Reduces manual review workload
• Scales to support large user bases

Comprehensive Coverage Detects both known and novel CSAM across multiple content types.

• Image, video, and text moderation
• Multi-method detection for robust protection