Trust & security
This page is maintained by My Life's Vault. It explains the controls that are live in the platform today and how we keep important actions auditable — so you can decide with eyes open.
Platform
The controls below describe how the platform is built and operated today. We are not SOC 2 Type II or ISO 27001 certified as yet, and we do not claim to be. We have designed and operate My Life's Vault against many of the security, privacy and operational principles that underpin these frameworks, including the GDPR regulatory framework. Rather than asking members, partners or advisors to rely on badges or marketing claims, we publish the controls and practices that are actually in place so you can judge our approach for yourself.
All traffic is served over modern TLS with HSTS. No sensitive data ever travels in URL query strings, error pages or analytics events.
Managed Postgres and object storage encrypt data at rest. Document and media buckets are private; downloads use short-lived signed URLs, never public links.
Each member-data table enforces row-level security scoped to the authenticated user. The privileged service role is restricted to verified server functions and signed webhooks — never the browser.
Workspaces are isolated by region and tenant. Production member data is not accessed for ad-hoc analytics; staff touch it only for named maintenance migrations or incident response.
Application
Strict TypeScript, ESLint and dependency scanning run in CI. Security scanner output is reviewed before each publish.
Authentication is enforced server-side, not by hiding UI. Sensitive actions — handover, exports, email or phone change, account deletion — require re-authentication, configurable in Settings → Security.
Email + password sign-in checks against the HIBP breached-password list. Google sign-in and MFA challenges are available, and an auto-lock can be enabled on the device.
Every inbound webhook is signature-verified before processing. Every server function validates its input with a schema — we don't trust the client.
Personal information
We treat the vault's contents — names, dates, family relationships, beneficiaries, health and care preferences, will and estate content, financial references, and identity documents you choose to upload — as sensitive personal information from the moment they're captured.
We capture only the field the member is filling in. Nothing is pre-filled from external profiling, ad networks or data brokers. We do not OCR uploaded identity documents into searchable fields by default.
ARIA will not ask for and will not store passwords, full card or account numbers, CVV, PIN, full national-ID numbers (SSN, NRIC, Emirates ID, passport), MFA codes, recovery seed phrases or private keys. If they're pasted in, ARIA asks the member to remove them.
Member data lives in the regional workspace's database with encryption at rest and row-level security scoped to your account. Sensitive free-text fields use field-level encryption where designed for it.
Your personal information is never sold, never used for advertising, and never used to train external AI models. Advisor and beneficiary portals see only the slice you've explicitly configured for them.
Health information, religious preferences (including Faraid / shariah choices) and family-structure data are treated as special-category data — extra access scoping, used only to serve your own plan.
You can export your vault and audit log at any time, and request deletion via privacy@mylifesvault.com. Signed wills are kept under legal-hold; partner-provisioned seats follow a 60-day offboarding window before purge.
Designed against
GDPR is a regulation, not a certification — no product can claim to be 'GDPR certified.' What we can show is how the platform is designed against its principles, with equivalents under UAE PDPL, Singapore PDPA and US state privacy laws.
We explain in plain language what each section of the vault is for, why we need a field, and who can see it. Lawful bases (contract, legitimate interests, consent) are never bundled together.
Vault data is used to serve your plan and the people you've nominated — not for unrelated analytics, profiling or marketing. We don't ask for fields we don't need.
Members can correct any record at any time. Signed wills are immutable by design; corrections produce a new witnessed version. Backups follow the managed database's retention.
Data residency follows the regional workspace you signed into. Sub-processor categories and the current list are available from the concierge as part of the DPA.
Access, rectification, restriction, objection, portability and deletion are supported. Self-service export covers vault data and the audit log; concierge handles the rest within statutory timeframes.
Security incidents are concierge-handled with a 24-hour acknowledgement target and a 72-hour authority-notification target where applicable. Members are notified when their personal information is materially affected.
Operations
Handover, exports, deletion and advisor access are member-initiated. A human concierge handles incidents, disclosures and complex requests — not a chatbot.
The managed database supports point-in-time recovery. Restore drills are exercised before major releases so backups are known-good, not assumed-good.
Members can export their own audit log as CSV at any time. The log is retained for accountability and to support the executor or advisor if access is ever questioned.
Researchers can report findings to security@mylifesvault.com. We acknowledge within one business day and credit responsible disclosure when fixes ship.
Today
We treat every record as though a family one day will rely on it during illness, incapacity or loss. The controls below are live in the platform today.
All traffic is served over TLS. Stored data, files and backups are encrypted at rest by our cloud infrastructure.
Every vault record is scoped by ownership and explicit permission. Trusted people see only what you have chosen to share with them, and only at the moment you allow.
Family members, advisors and trusted contacts live in a single People & Relationships registry. Roles and permissions are tracked per person — no duplicate identities, no orphaned access.
We collect only what is needed to make the vault useful. Sensitive content is never used to train external models, and we never sell or share your data with third parties for marketing.
Auditability
When something important happens — a will is signed, an executor accepts, a death is reported — we capture who did what, when, and from where.
Once a will is signed it becomes a sealed version. Edits create a new version that must be re-witnessed. Every prior version is preserved for the executor and for the courts.
Each witness signs through a unique link. We capture timestamp and IP, and attach a tamper-evident certificate to the final PDF.
Nominees confirm, decline or ask questions through their own secure link. Their response — and the testator's notification — is logged on the person record.
Every check-in, escalation step and trusted-contact notification is recorded so the family can see exactly what happened and when.
Handover
The platform supports temporary incapacity, long-term incapacity and death scenarios — with different access rules for each, and the ability to roll back where appropriate.
Short-term handovers (a hospital stay, travel emergency) can be granted, time-boxed and withdrawn. The vault never discloses more than you intended.
Death-event handover only unlocks the scope you defined — beneficiary information, executor instructions, the records each trusted person was meant to receive.
When a death or incapacity is reported, a human concierge acknowledges within 24 hours and guides the next steps. Not a chatbot.
Sub-processors
We rely on a small, deliberate set of vendors. The current named list is shared with members on request and forms part of the Data Processing Addendum.
Hosts the regional workspace databases and private file storage with encryption at rest, row-level security and point-in-time recovery.
Delivers concierge, witness, executor and notification emails. Sensitive content is redacted from email bodies wherever possible.
Routes ARIA's prompts to model providers under contracts that prohibit training on member data. ARIA's guardrails strip secrets before any prompt leaves the platform.
A PCI-handling payments provider processes subscriptions (we never store card details), and a monitoring provider tracks uptime and errors without ingesting member PII.
Working towards
These are intent, not commitments — no dates, no certification claims. We'll publish updates as items move from planned to in progress to live.
Planned. An independent third party will test the published platform; the executive summary will be available under NDA to enterprise partners.
In progress. A public, versioned sub-processor register with change-notification for partners.
Planned. A readiness assessment against the SOC 2 Trust Services Criteria to prioritise the highest-value controls before any formal audit.
Planned. A gap assessment against ISO/IEC 27001 to scope our information security management system.
Planned. A formal programme with safe-harbour terms and clear scope, building on the coordinated disclosure channel already open at security@mylifesvault.com.
Our concierge team responds within one business day. For security disclosures, please mark your email Security.