Skip to main content

Trust

How Draftless handles your work

What we do with your data, who processes it, and the safeguards built into every paid run. Each statement here is checked against our code by an automated test.

Data use and training

  • Your content is not used for training unless you switch it on. Our servers treat anything other than an explicit opt-in as no.
  • In a team workspace the policy Never overrides any member’s choice, and a workspace without a policy is treated as Never.
  • A model whose provider may retain or learn from inputs is labelled, and runs only after you confirm it once. The server refuses the run until you do.

Privacy policyData Processing AgreementDelete your account

Security practices

  • Row-level security is enabled on every database table we create, so the database itself checks who may read or change each row.
  • Paid runs accept an idempotency key: a retry with the same key returns the first result instead of charging again. Each card payment grants its credits exactly once.
  • If a paid run fails after credits were taken, the credits go back to your balance automatically, once.
  • Cloning a voice, animating a portrait and swapping a face each require an explicit consent affirmation, checked on our server before anything is charged. The wording and its version are written by the server, not by the app.
  • Images saved as Elements, the reusable references, are screened by a vision model, and any image showing a person who may be under 18 is rejected. AI Influencer briefs pass a separate child-safety check that refuses whenever it cannot give a clear answer.

Content provenance

  • Every cloned-voice result carries a plain-language notice that it is AI-generated speech, and we keep a provenance record for it in our database.
  • An animated-portrait result comes back with the same kind of AI-generated notice in its result data.
  • We do not yet embed a watermark or a signed content credential in the files themselves. The notice and record are what we provide today.

Sub-processors

Infrastructure providers, and the AI model providers that receive the prompt and input media of a run you start. Training posture is as we read each provider’s terms.

Infrastructure

  • Managed Postgres / backend hostaccount, application and billing data. Database, auth, storage and server functions. Encrypted at rest and in transit; region per plan.
  • Object storagegenerated and uploaded media, for persistence and delivery. Versioned and access-controlled.
  • Error monitoring (Sentry)error events, request identifiers and limited context, for reliability and debugging, plus session replays for a sample of sessions and for sessions where an error occurs. Personal data is scrubbed from breadcrumbs where possible.
  • Email deliveryemail address and message content, for transactional email: authentication, receipts and notifications.
  • Stripebilling identifiers and payment metadata, for payments and subscriptions. Stripe is the PCI merchant of record; card data never reaches our servers.
  • Google Play Billingpurchases in the Android app. Google receives the payment details; we receive a purchase token and order ID only.
  • ElevenLabsvoice samples you upload for voice cloning, to create the cloned voice.

AI model providers

ProviderTraining on your inputs
OpenAIDoes not train on your inputs
BytePlusDoes not train on your inputs
Lovable AI gatewayNot confirmed — terms unclear
fal.aiUses only anonymised usage data
KlingNot confirmed — terms unclear
RunwayNot confirmed — terms unclear
ReplicateNot confirmed — terms unclear
ElevenLabsNot confirmed — terms unclear
HiggsfieldMay retain or learn from inputs
IdeogramNot confirmed — terms unclear
RecraftNot confirmed — terms unclear
Stability AINot confirmed — terms unclear
Luma AINot confirmed — terms unclear
fluxapi.aiNot confirmed — terms unclear
sunoapi.orgNot confirmed — terms unclear
Memories.aiNot confirmed — terms unclear

Full sub-processor listData Processing Agreement

Report a security issue

Found a vulnerability? Email us with the steps to reproduce it. Please give us a chance to fix it before you share it publicly. For questions about your data, write to the privacy address.

Help centre