The Roadmap.
A public planning document for a facility that's evolving from an education platform into trust infrastructure for a world where humans and AI create together.
S.M.I. exists to establish trust, ownership, and authenticity in human–AI co-creation. We help creators verify who they are, preserve the provenance of what they make, protect their IP through documentation and evidence, and improve how their work is discovered by AI-driven search. We don't promise to stop deepfakes, block bots, or guarantee authenticity. We build the tools, records, and signals that let humans decide — with better evidence in front of them.
Four phases. No promised dates.
We move by phase, not by calendar. Each phase gets shipped when it's honest to ship.
The Academy
- AI education and synthetic-media literacy for creators, students, and investigators.
- Training simulations and playable field drills that teach detection thinking.
- The Intelligence Archive — a public knowledge base on generative media.
- Community foundations: Case Files, Learning Pathways, Investigator Desk.
The Registry
- Verifiable creator identity records tied to an investigator profile.
- Asset registration with timestamps and version / edit history.
- Voluntary AI-disclosure labels: Human-created · AI-assisted · AI-generated · Hybrid — mapped to IPTC DigitalSourceType where the vocabulary matches.
- Cryptographic anchoring — every record committed to a tamper-evident SHA-256 hash chain, with post-quantum signatures on the roadmap.
- One-click evidence packages: report, claims, license ledger, disclosure, and hash-chain proofs in a single export.
- Shipped ahead of schedule: a License Market where verified work is licensed on the record — every grant lands on the asset's public ledger and travels with the file as a C2PA assertion.
Verification & Monitoring
- Deepfake and synthetic-media indicators with confidence scoring — never absolute claims.
- Metadata inspection and provenance verification for uploaded material.
- Assistive analysis reports for creators, journalists, and rights holders.
- Monitoring that surfaces potential matches to registered works for human review.
The Ecosystem
Trust infrastructure, in seven pieces.
Each pillar is a load-bearing wall of what S.M.I. is becoming. None of them promise certainty. All of them promise transparency.
Digital Identity
Verifiable creator identities so authorship travels with the work — even after it leaves the creator's hands.
Digital Provenance
A chain of custody: timestamps, version and edit history, ownership chain, licensing history, and disclosed AI involvement.
IP Documentation
Organized evidence for copyrighted works. Records that support protection and complement — never replace — official copyright registration.
Content Verification
Assistive analysis with confidence scoring. Signals for humans to weigh, not verdicts to accept.
Intelligent Monitoring
Notification when public content potentially matches registered works — with evidence packaged for creator review.
GEO — Generative Engine Optimization
Helping organizations be accurately understood, referenced, and attributed by AI systems: structured metadata, knowledge graphs, provenance signals, and machine-readable citations.
Quantum-Ready Security
The Registry anchors every record in a tamper-evident SHA-256 hash chain today. The roadmap moves toward post-quantum signature schemes — so ownership proofs written now remain verifiable after quantum computing breaks classical assumptions.
Creator Control
Machine-readable usage policies, crawler preference management, watermarking and provenance metadata, licensing APIs — transparency and accountability, with an honest acknowledgment that no technical system can block every bot.
Trust requires accountability. Including ours.
A system built to establish trust has to be trustworthy first. These are the standing commitments S.M.I. holds itself to — the protocols we ship every phase against.
Automated analysis informs. Humans decide. S.M.I. tools produce assessments and evidence — never verdicts.
Confidence scores over certainty claims. Published methodology principles. Honest labeling of what is simulated versus real analysis inside the facility.
Creators choose what to register and what to disclose. Nothing about the Registry works by default — it works by consent.
S.M.I. operates inside existing legal frameworks — copyright, DMCA, licensing — and generates evidence that supports them rather than claiming to replace them.
S.M.I. labels its own AI-generated content, including AI art in the Vault. If we ask creators to disclose, we disclose.
This roadmap represents direction and intent — not guarantees, not fixed timelines, and not a finished map. Scope, sequencing, and naming will evolve as the field evolves and as we learn what actually helps creators. Everything on this page is subject to change as S.M.I. moves through each phase.
Come with us — or watch us build.
Reserve a place on the roadmap waitlist and we'll signal when new phases go live. Or head straight into the Academy — it's already open.