Examine the artifact behind the claim.
Photographs, voicemails, documents, and video can be generated or altered with widely available tools. When authenticity is disputed, the examination needs to address how the file was created, what it contains, and which conclusions the available evidence supports.
We examine the artifact and its available context, document the methods used, and prepare findings for counsel, expert declaration, and courtroom testimony.
Is the artifact authentic?
We analyze images, audio, video, and documents for indicators of synthetic generation or later alteration. The work can include compression signatures, metadata anomalies, acoustic artifacts, lighting and shadow inconsistencies, and frequency-domain patterns associated with generative tools.
An indicator is assessed in context. We document what we observe, the methodology, and the limitations of the available file rather than treating a detector's score as a conclusion.
What can be said about its source?
Attribution is a separate question from authenticity. For potentially generated text or images, we examine available provenance data, stylometric markers, prompt-related artifacts, and patterns across source material.
The available evidence may support a limited conclusion or leave attribution unresolved. Our reporting identifies those limits and the assumptions on which an opinion depends.
A human expert directs the analysis.
We use AI tools within a forensic workflow, with a human expert responsible for the analysis and conclusions. The expert documents the methodology and can explain it at deposition and trial.
The same principle guides our social media and OSINT monitoring: expert direction, validation of findings, and preservation suitable for evidentiary use.
Scoping the engagement.
Describe the image, recording, video, or document, how it was obtained, and the authenticity question in dispute. Include whether the original file and any surrounding records are available.