Truth infrastructure for the AI era
Dispel the fog.
See what's real.
Dispel inspects suspicious text, images, voice notes, music, video, screenshots, links, and digital artifacts using provenance, synthetic indicators, contradiction checks, and calibrated trust scoring.
PROVENANCE
SCAM RISK
CLAIMS
MANIPULATION
LIVE SIGNAL RESOLUTION
The cursor cuts through the fog. The system resolves what can be defended, and leaves weak evidence marked as unknown.
TRUST IS NOW A TECHNICAL PROBLEM
Ambiguity becomes structure. Structure becomes evidence. Evidence becomes action.
Origin
Manipulation
Provenance
Scam Risk
Claims
Action
FOG TO CLARITY
Suspicion enters as noise. It exits as a readable verdict.
Inspect the suspicious surface
Emails, screenshots, voice notes, images, clips, listings, pitches, invoices, links, and messages all enter the same trust layer.
Cross-map evidence categories
Dispel correlates provenance, persuasion markers, synthetic signatures, claim consistency, and source mismatch without pretending weak evidence is certainty.
Condense complexity into action
The result is a calibrated verdict, subscore matrix, evidence list, uncertainty framing, and recommended next step.
CALIBRATED OUTPUT
Dispel does not read minds. It reads signals.
Outputs stay premium, sharp, and honest: likely real, mixed, suspicious, high risk, or unknown when the evidence does not support a stronger claim.
SUSPICIOUS
TRUST SCORE 34
SCAM RISK
78
PROVENANCE
12
CLAIMS
29
AI LIKELIHOOD
61
- Urgency language and payment pressure detected
- Source identity unsupported by verifiable origin
- Claim set exceeds available evidence
- Recommended action: do not pay before independent verification
MULTI-MODAL DETECTION WALL
One trust layer across every suspicious format.
Scam rhetoric, pressure, impersonation, contradiction
Metadata, provenance gaps, manipulation and synthetic markers
Container signals, frame evidence, transcript-linked risk cues
Transcript-first scam signals, synthetic speech capability when available
Synthetic provenance cues without fake ownership claims
Compact verdict receipts designed to be shared and understood fast
SCAM THEATER
Watch deception collapse under evidence.
CASE 01
Fake recruiter email
An urgent “you’re hired” message requests identity documents and payment for equipment before any verified interview process exists.
- Authority impersonation structure
- Payment pressure before onboarding proof
- Domain and source mismatch
- Action: verify via public company channel
CASE 02
Cloned emergency voice note
A distressed voice demands immediate money transfer, but origin markers and speaker authenticity proof are absent.
- Urgency spike and emotional coercion
- Identity cannot be confirmed from available evidence
- Provenance status incomplete
- Action: verify by independent contact path
CASE 03
Manipulated marketplace listing
Product imagery looks persuasive, but the seller pushes off-platform payment and details shift under basic scrutiny.
- Manipulation risk signals in media
- Pressure to abandon platform safeguards
- Claim inconsistency across messages
- Action: keep payment inside trusted rails
CASE 04
Crypto bait ad
Guaranteed upside, celebrity-style credibility, synthetic promotional tone, and no verifiable provenance.
- Unsupported extraordinary claims
- Synthetic persuasion markers
- Origin confidence remains weak
- Action: do not send funds from ad copy alone
EVIDENCE-FIRST CORE
The product philosophy is simple: signal beats spectacle.
EVIDENCE
CALIBRATED
Metadata & provenance
When origin markers are present, Dispel uses them. When they are missing, it says so directly.
Synthetic indicators
Provider-backed signals and local heuristics are separated clearly so confidence does not get inflated by branding.
Scam structures
Urgency, coercion, authority theater, payment pressure, and manipulated context are surfaced in plain language.
Claim consistency
Unsupported, contradictory, or unverifiable claims are flagged before the user trusts the source or takes action.
PRODUCT SURFACES
Designed like an instrument, not a dashboard template.
Verdict-first investigations with subscores and operator-grade evidence.
Saved investigations with continuity across media types.
Shareable truth receipts that are clear in under five seconds.
Programmatic trust scoring for platforms, moderation tools, creator protection, and trust & safety systems.
PRICING
Consumer clarity. Professional protection. Platform infrastructure.
$0
Core checks for suspicious text and image inputs.
$19
Higher-volume personal protection with richer reports and media coverage.
$49
API access and operational trust tooling for creators, operators, and teams.
FAQ
Clarity over hype.
Can Dispel prove something is fake?
No. It evaluates available signals and explains what is defensible, what is weak, and what remains unknown.
Is this a lie detector?
No. Dispel analyzes evidence, provenance, synthetic indicators, scam patterns, and claim consistency. It does not claim mind reading.
Why does “Unknown” matter?
Because honest uncertainty is more useful than fake certainty. Unknown means the evidence is not strong enough to support a stronger call.
FINAL CONVERSION
Truth, before trust.
For suspicious media, uncertain messages, and high-stakes digital decisions, run the evidence first.