Synthetic
Forensics
Forensic methods for synthetic media, AI-generated artifacts, provenance, authenticity, and human-AI interaction.
State what the evidence supports.
What is synthetic forensics?
Synthetic forensics is the systematic identification, preservation, examination, reconstruction, and interpretation of digital evidence involving artificial intelligence systems, synthetic media, AI-generated artifacts, and human-AI interactions.
Evidence may include conversations, application storage, generated files, tool records, and provenance. Start by scoping the question and preserving available sources. A detector score alone does not reconstruct a workflow.
Read the definition, scope, and limitations
Core research areas.
Ten connected areas, from media and source attribution to human interpretation and evidentiary integrity. Each begins with a question that evidence can test.
Synthetic Media Forensics
Examination of generated or altered image, video, audio, and multimodal content.
Read referenceReferenceAI-Generated Artifacts
Generated files and residual application, device, and agent records surrounding AI use.
Read referenceReferenceProvenance & Authenticity
Origin, lineage, transformations, and the support for claims about what an object is.
Read referenceReferenceDetection & Attribution
Testing synthetic-content signals and associations with a source, model, workflow, or actor.
Read referenceReferenceEvidence Reconstruction
Recovering supported event sequences and source relationships from incomplete digital records.
Read referenceReferenceHuman-AI Interaction
Records and interpretation of exchanges between people and AI-enabled systems.
Read referenceReferenceSynthetic Memory & Source Monitoring
Research on how people distinguish experienced, communicated, imagined, and synthetic information.
Read referenceReferenceChain of Custody & Evidence Integrity
Preserving bytes, handling history, versions, and traceability of original and derived evidence.
Read referenceReferenceAI Anti-Forensics
Defensive study of interference with the availability or interpretation of AI-related evidence.
Read referenceReferenceLegal & Evidentiary Issues
Authentication, method reliability, disclosure, and the limits of forensic interpretation.
Read referenceUnderstand the terms behind the evidence.
Read 300 plain-language definitions with practical explanations, related concepts, and primary references. Explore evidence handling, synthetic media, provenance, AI conversations, agents, models, and identity. Filter by a word or evidence field, or follow a stable link to the term you need.
Explore the 300-term glossaryMethods turn terminology into an examination. Research connects the questions across disciplines.
How we evaluate synthetic evidence.
Preserve the sources, establish their history, inspect metadata and system records, analyze content, and test competing explanations. Seek independent corroboration and state what remains unresolved.
Follow the layered evaluation methodA detector output is a signal within an examination. It needs relevant validation and an explained relationship to the claim.
Start with the evidence.
An output is only one source. Conversations, application state, tool actions, and provenance can explain how it came to exist.
Synthetic Media
Examine the media object, its transformations, and the records behind it.
Read reference02AI Artifacts
Recover meaning from application state, storage, logs, and exports.
Read reference03Interaction Forensics
Reconstruct exchanges, context, revisions, and downstream actions.
Read reference06Provenance
Verify recorded history, signed assertions, bindings, and trust.
Read referenceAI Agent Forensics
Follow delegated actions from task initiation to external effects.
Read reference05Model Forensics
Assess model lineage, configuration, identity claims, and integrity.
Read reference07Synthetic Identity Forensics
Examine represented identities, account traces, and coordination.
Read referenceAI Agent Forensics
An agent says it changed a file or sent a report. What evidence supports that claim? Follow the task, tool calls, permissions, and records from the affected system to establish what happened.
Read the AI Agent Forensics guideFrom an interaction to evidence.
Not every AI platform produces every artifact type. Availability depends on version, settings, retention, access, and collection.
- 01Human
- 02AI System
- 03Response / Tool Action
- 04Local / Cloud / Application Artifacts
- 05Transformation / Export
- 06Forensic Acquisition
A framework for a defensible examination.
The Synthetic Forensics Investigation Framework organizes ten stages from identification to reporting. It makes validation and uncertainty visible throughout the work.
Explore SFIFPublished by SyntheticForensics.org. Project-authored methodology with explicit scope and limits.
- 01Identify
- 02Preserve
- 03Acquire
- 04Extract
- 05Reconstruct
- 06Correlate
- 07Attribute
- 08Validate
- 09Interpret
- 10Report
Build a traceable conclusion.
Evidence Taxonomy
Nine evidence classes, with persistence and corroboration cautions.
Read reference02Methods
From scoping and acquisition to interpretation and reporting.
Read reference03Standards
Primary specifications and guidance, with scope made explicit.
Read reference04Glossary
Stable definitions for the language of AI-related evidence.
Read reference05Artifact Reports
A reusable format for versioned, testable artifact observations.
Read reference06Case Studies
Five hypothetical examinations, including alternative explanations.
Read referenceTools and projects.
Explore the investigation framework, report specification, and reference downloads. The tool catalog describes software by purpose, evidence types, and validation conditions. Use the source archive and GitHub gateway for technical inspection.
View current projectsLearn, question, contribute.
Use the reference now and contribute a clear question, reproducible method, or useful explanation.
Courses and study path
Begin with the free reference. Explore courses, lessons, and practical learning resources.
Read referenceDiscussCommunity
Help improve methods and source interpretation. Contribute questions, corrections, and careful technical discussion.
Read referenceDevelopRoadmap
Open research questions and the sequence for the platform’s next capabilities.
Read referenceParticipateContribute
Prepare a reproducible observation, correction, or bounded research proposal.
Read referenceRelated projects.
Resources concerning digital evidence and the human interpretation of synthetic information.
Synthetic Imprinting
The human and cognitive side of synthetic information: source monitoring, memory, and interpretation. Cognitive inquiry and digital forensics are distinct disciplines.
Anti-Forensics.com
An adjacent resource on digital forensics and anti-forensics, including persistence, preservation, and the limits of artifact interpretation.
Related-project links do not imply institutional affiliation, endorsement, or validation of this reference.