Subprocessors
Last updated: May 4, 2026
Lantern uses the following third-party service providers (known as "subprocessors") to operate the Service. Each provider processes a specific, limited subset of user data as described below. This page is the authoritative list.
Current subprocessors
Resend
- Provider
- Resend, Inc. (Delaware, USA)
- Service
- Transactional email delivery
- Data processed
- Your email address, your display name, and the content of transactional emails we send you (account verification links and password reset links)
- Purpose
- Sending account verification emails and password reset emails
- Data processing agreement
- resend.com/legal/dpa
Google (Sign in with Google)
- Provider
- Google LLC (Delaware, USA)
- Service
- OAuth identity provider for the optional "Continue with Google" sign-in flow
- Data processed
- Only the Google profile fields you consent to at sign-in: your email address, your display name, your profile picture URL, and Google's opaque stable user ID (the "sub" claim). No Google data is processed if you do not use the "Continue with Google" button.
- Purpose
- Verifying you are the owner of the Google account so we can sign you in to Lantern. We do not read any other Google services (no Drive, Gmail, Calendar, Contacts, Photos, etc.).
- Data processing agreement
- cloud.google.com/terms/data-processing-addendum
Modal
- Provider
- Modal Labs, Inc. (Delaware, USA)
- Service
- Serverless GPU inference for the neural matcher (SSCD, CLIP, DINOv3)
- Data processed
- Image bytes you upload at registration or verification are forwarded to Modal for embedding computation. Modal returns a numerical embedding vector (hundreds to thousands of floats) to Lantern; the embedding is stored in our database. Image bytes themselves are processed in a transient container on Modal and not retained by Modal or by Lantern after processing. No account identifiers or personal metadata (email, display name, wallet address) are sent to Modal.
- Purpose
- Computing neural embeddings that power Lantern's detection of cropped, flipped, recompressed, and filtered copies of registered works. These embeddings are the technical basis for the accuracy upgrade shipped in Protection Phase 1.
- Data processing agreement
- modal.com/legal/dpa
- Model attribution
- Lantern's neural matcher is built with DINOv3, released by Meta AI Research under the DINOv3 license. See github.com/facebookresearch/dinov3 for the model and license.
DigitalOcean
- Provider
- DigitalOcean, LLC (New York, USA)
- Service
- Cloud hosting for the Lantern application server and managed PostgreSQL database
- Data processed
- All account data we store (email, display name, encrypted wallet key, registration records, dispute history) resides on DigitalOcean infrastructure. DigitalOcean does not have application-level access to this data; they operate the underlying compute and storage.
- Purpose
- Running the Lantern API and storing account data
- Data processing agreement
- digitalocean.com/legal/data-processing-agreement
Alchemy
- Provider
- Alchemy Insights, Inc. (San Francisco, USA)
- Service
- Ethereum / Base blockchain RPC access
- Data processed
- When a registration is written to the Base blockchain, Lantern submits the transaction (content hash, metadata URI, creator wallet address) via Alchemy's RPC endpoint. No email address, display name, or other personal identifiers cross this boundary; only the pseudonymous wallet address and the content hash, which are by design public on the blockchain anyway.
- Purpose
- Relaying on-chain registration transactions to the Base network
- Data processing agreement
- alchemy.com/policies/data-processing-addendum
Changes to this list
When we add or remove a subprocessor, we will update this page and notify registered users by email at least 30 days in advance of the change, except in cases where immediate replacement is required to keep the Service operating.
Questions
Email contact@lantern-us.com with any questions about how your data is processed.