In CSAM and safeguarding investigations, one of the most demanding aspects of the work is not locating material, but repeatedly viewing and describing it in detail. Investigators are often required to document disturbing images and video frames multiple times across reports, reviews, and case stages.
While this work is unavoidable, repeated exposure does not improve investigative outcomes. It increases cognitive and emotional load, slows progress, and places sustained pressure on investigator well-being.
Reducing unnecessary exposure is not optional — it is an operational responsibility.
To address this, a capability was developed and used operationally inside S21 VisionX to generate structured, consistent descriptions of images and video frames, reducing the need for repeated manual interpretation while preserving investigative detail.
Following daily operational use and demand from investigative teams, Semantics 21 has now made this capability available as a standalone tool: S21 AI Describe.
The tool generates clear, structured descriptions of visual content, allowing investigators to understand and document material without prolonged or repeated viewing. Outputs preserve context and detail and are routinely used to support further searches, analytical work, and investigative decision-making.
AI Describe operates fully offline and does not share sensitive material with third parties.
It is designed to integrate into existing investigative workflows, without replacing tools, retraining staff, or changing how teams operate.
Semantics 21 has positioned AI Describe as part of a broader shift toward sustainable investigative practice, where reducing unnecessary exposure is recognised as a baseline requirement of modern digital investigations.
More information and demonstrations are available at semantics21.com/s21-ai-describe
