Vallum Protocol: verifiable trust for AI systems and enterprise operations.
Vallum is independent trust and evidence infrastructure for organizations operating AI agents, automated systems, APIs, enterprise applications, and high-trust business workflows.
It establishes scoped authorization, verifies evidence integrity, preserves cryptographic provenance, issues verifiable certificates, and produces defensible evidence packages.
Vallum integrates with the systems an organization already operates. It does not replace endpoint security, identity management, network security, cloud security, governance, or legal review.
What is the Vallum Protocol?
Vallum defines how protected systems prove authority, evidence integrity, provenance, and accountability.
Before a protected operation proceeds, Vallum can require scoped authorization. After the operation, it can verify the resulting evidence, validate signed artifacts, preserve the chain of custody, and produce independently verifiable records.
Vallum is designed for AI agents, automated services, enterprise applications, APIs, local nodes, and cross-system workflows where ordinary logs are not enough.
Vallum does not give an AI system unlimited authority. It establishes evidence-backed limits on what the system may do and creates proof of what occurred.
Core safety commitments
- No human coding enforcement. Vallum must never restrict a person's rights-tier access because of a barcode, account token, implant, reputation score, behavioral score, identity profile, or automated risk classification.
- No autonomous external escalation. Vallum does not contact authorities, file external reports, or initiate irreversible actions about a person without explicit human authorization and an applicable lawful process.
- No rights restriction through scoring. Vallum does not create social-credit classes or use automated scoring to remove fundamental user capabilities.
- Evidence integrity. Issued certificates, signed artifacts, evidence manifests, and chain-of-custody records are protected against silent alteration. Revocation, expiration, and renewal create auditable lifecycle events rather than rewriting history.
- Tenant-scoped authority. Protected cross-tenant actions must fail closed. Authorization for one tenant cannot be reused to access another tenant's protected records.
- Full trust-path auditability. Security-relevant authorization decisions, evidence-validation results, signature checks, certificate lifecycle events, and attempted boundary violations are recorded for authorized review.
The Titan Stack: Vallum's proof and trust foundation
Vallum defines the trust model. Four specialized Titan systems provide authorization, verification, provenance, and adoption capabilities.
Titan Handshake
Establishes scoped authority before protected systems exchange data or perform high-trust actions.
- Tenant-scoped authorization receipts
- Time-bounded and purpose-bounded authority
- Replay and scope validation
- Fail-closed protected-action checks
Titan Verify
Validates evidence packages, manifests, schemas, hashes, and chain-of-custody records.
- SHA-256 evidence and manifest verification
- Versioned structural validation
- Tamper and corruption detection
- Offline and auditor-oriented verification
Titan Sign
Provides cryptographic provenance for reports, certificates, manifests, and final evidence artifacts.
- Ed25519 digital signatures
- Signed-root and manifest verification
- Provenance metadata
- Tamper-evident issued artifacts
Titan Forge
Provides SDKs, command-line tools, templates, fixtures, and integration guidance for teams adopting Vallum and the Titan protocols.
- SDK and CLI adoption tools
- Templates and conformance fixtures
- Local readiness checks
- Integration and developer documentation
Titan Forge may report local readiness, but it cannot issue an official VERIFIED or CERTIFIED result. Official certification belongs to Verified Ops.
Verified Ops: certification and evidence delivery
Verified Ops is the commercial trust-authority layer built on the Titan stack. It coordinates protected authorization, evidence verification, signed-report validation, entitlement checks, certificate issuance, public verification, revocation, renewal, evidence-package export, human-readable evidence binders, and signed evidence delivery.
Payment never creates trust. A paid entitlement only allows a qualifying result to receive CERTIFIED status after the required authority, evidence, verification, and signature checks pass.
The submitted evidence passed the applicable technical verification path but did not receive paid certification status.
The required authority, evidence, verification, signature, and active certification-entitlement checks passed.
Forge readiness, payment status, or a local test result can never independently produce CERTIFIED status.
What Vallum is designed to prevent
These are technical and architectural boundaries, not vague marketing promises.
How Vallum works
Protected actions can be authorized
Titan Handshake establishes which system, service, agent, or operator is authorized to perform a protected action, for which tenant, for what purpose, and during what time window.
Evidence can be verified
Titan Verify validates evidence packages, manifests, hashes, schemas, and chain-of-custody records.
Important artifacts can be signed
Titan Sign provides cryptographic provenance for certificates, reports, manifests, evidence packages, and final delivery artifacts. Doctrine: sign the root, not the wrapper.
Organizations can integrate safely
Titan Forge provides SDKs, command-line tools, templates, fixtures, and readiness checks for connecting existing systems to Vallum.
Certification and evidence delivery
Verified Ops coordinates certificate issuance, public verification, evidence-package export, evidence binders, signed delivery archives, revocation, renewal, and lifecycle records.
Public verification stays minimal
Public verification can confirm the status and authenticity of an issued certificate without exposing tenant-private evidence, internal billing records, secrets, or unrelated customer data.
Evidence packages preserve depth
Customers can receive structured evidence packages and human-readable binders containing certificate records, verification summaries, signature results, chain-of-custody timelines, hash manifests, and public-verification references. Full machine evidence remains in the protected evidence archive rather than being dumped into the public verification page or the main binder.
Privacy by architecture
Vallum can integrate with separately deployed privacy-smoothing controls designed to reduce unnecessary metadata leakage from supported automated operations.
The Nimble White Noise Agent is a separate privacy module. It may emit only clearly classified synthetic operational noise and must never create fake users, customers, reviews, leads, traffic, engagement, or compliance evidence.
More than “we take security seriously”
Where Vallum fits
Vallum can support protected workflows across:
Vallum surrounds high-trust operations with authorization, verification, provenance, certification, and evidence controls without becoming the application itself.
Vallum does not provide legal advice
Vallum is technical trust, evidence, and safety infrastructure. It does not provide legal advice, legal representation, regulatory approval, or automatic legal compliance certification.
Certificates and evidence packages attest only to the technical evidence and verification scope stated in the artifact. Organizations should consult qualified legal, regulatory, security, audit, and compliance professionals for obligations specific to their industry, jurisdiction, or contract.
Trust built into the system—not added after an incident.
Use Vallum to authorize protected actions, verify evidence, preserve provenance, issue certificates, and deliver independently verifiable evidence packages.