AI-generated media & deepfake detection
Assess synthetic-image indicators and video deepfake cues using model- and artefact-based signals.
Multi-signal media authentication
TattvAI
Correlate the signals. Document the decision.
An evidence-oriented platform combining AI-media detection, manipulation forensics, provenance checks and analyst workflows. TattvAI correlates independent signals to support triage, review and documented investigative decisions.
Investigation canvas
Multiple independent signals. One accountable assessment.
TattvAI brings model, pixel, sensor, compression, provenance and temporal signals together for explainable analyst review.
10Capability domains
48Configurable analysis filters
MultiImage + video support
SHA-256Specimen identity binding
Manipulation, clone & PRNU analysis
Surface copy-move regions, local inconsistencies and sensor-noise cues for tampering triage.
JPEG & recompression forensics
Examine quantisation, double compression, ghosting, blocking and recompression anomalies.
Metadata, geolocation & provenance
Review EXIF, timestamps, device data, GPS context and provenance consistency.
Frame-by-frame video analysis
Navigate timelines, inspect key frames and assess anomalies at individual-frame level.
Analyst annotations & investigation tools
Mark regions of interest, add notes and bookmarks, compare evidence and lead a structured review.
Batch, history & project workflows
Organise cases, bulk processing, saved analyses, audit history and repeatable workflows.
All 48 configurable analysis filters
Combine, tune and compare forensic filters for the media type and case context.
Evidence reports & SHA-256 identity
Produce structured findings while binding every specimen to a SHA-256 cryptographic hash.
Investigation workflow & responsible use
Move from triage to report with explainable signals, stated limitations and analyst sign-off.
From specimen identity to evidence report.
Each investigation combines six signal families: AI/deepfake, pixel, sensor/PRNU, compression, provenance and temporal continuity.
Ingest & hash
Register the image or video and bind the specimen to a SHA-256 identity.
Extract signals
Run selected model, pixel, sensor, compression, provenance and temporal checks.
Correlate findings
Compare independent indicators instead of relying on a single score.
Analyst review
Interpret findings against case context, source evidence and known limitations.
Evidence report
Document the analysis, reasoning, limitations and analyst sign-off.
For investigations where authenticity cannot be assumed.
TattvAI supports evidence-led review across public, media, commercial and legal contexts.
Law enforcement & cybercrime
Assess questioned media through a structured, multi-signal forensic workflow.
Newsroom & media verification
Review provenance, manipulation and synthetic-media indicators before publication or reliance.
Insurance & claims review
Investigate submitted images and videos with repeatable analysis and specimen identity controls.
Brand, UGC & content integrity
Examine user-generated and brand-related media for manipulation and AI-generation signals.
Enterprise & legal investigations
Give legal, compliance and security teams documented findings backed by analyst review.
Every signal, visible in context.
Move from the investigation canvas into model, pixel, provenance and temporal views without losing specimen identity or review history.
Governance for defensible media review.
TattvAI keeps findings explainable, source-bound and subject to trained human interpretation.
Responsible use
No single filter determines authenticity.
TattvAI provides multi-signal decision support, not a standalone determination of authenticity. A trained analyst should interpret findings against case context and source evidence, state limitations, and document applicable legal, privacy and chain-of-custody controls.
Make visual authenticity reviewable.
Bring us your investigation workflow, media types and reporting expectations. We’ll show how TattvAI can fit the review process.