AI Use and Commercial Licensing
FormosanBank is available for noncommercial research, education, cultural work, and language revitalization, subject to each corpus's source terms and citation requirements.
Commercial AI use is not permitted without prior written permission from FormosanBank and any applicable rights holders.
What counts as Commercial AI Use?
Commercial AI Use means using FormosanBank materials, in whole or in part, for an AI system, dataset, benchmark, product, service, or internal workflow that is commercial, revenue-generating, investor-funded, product-development-oriented, or used to support commercial operations.
This includes:
training, pretraining, fine-tuning, distilling, or adapting models
creating embeddings, vector indexes, retrieval systems, or RAG systems
evaluating, benchmarking, red-teaming, or filtering AI systems
generating synthetic data, translations, speech, annotations, or derived corpora
packaging FormosanBank-derived data, features, model weights, adapters, outputs, or embeddings into a commercial product or service
Uses that usually do not require separate commercial permission
These uses are generally allowed if they follow the Terms of Use, cite FormosanBank and the relevant source corpora, preserve notices, and respect any stricter corpus-specific restrictions:
noncommercial academic research
classroom use
community language revitalization
cultural, archival, or documentation work
personal study
noncommercial computational experiments whose outputs are not sold, licensed, or deployed commercially
If a corpus page, source publication, XML metadata, dataset card, or contributor agreement has stricter terms, those stricter terms control.
Uses that require written permission
Do not use FormosanBank materials without written permission for:
commercial AI training, fine-tuning, evaluation, embeddings, RAG, or synthetic-data generation
commercial speech recognition, translation, text generation, search, analytics, or language-technology products
resale, relicensing, paid hosting, or paid API access to FormosanBank materials or derivatives
removing, weakening, or hiding source, attribution, license, provenance, or no-commercial-AI notices
using FormosanBank data in a way that conflicts with source-community expectations, contributor restrictions, privacy constraints, or source-corpus licenses
AI Use Addendum
The full AI Use Addendum is maintained with the public repository:
That addendum defines Commercial AI Use in more detail and states the remedies FormosanBank reserves for unauthorized commercial AI mining, including license-back terms for resulting models, weights, embeddings, datasets, products, and derivatives to the maximum extent permitted by law.
Requesting commercial permission
Before any commercial use, contact the FormosanBank team with:
the corpus or corpora you want to use
the organization requesting access
the intended product, service, model, or workflow
whether the materials will be used for training, fine-tuning, embeddings, RAG, evaluation, or synthetic data
whether outputs, models, embeddings, datasets, or derivatives will be shared, sold, hosted, or redistributed
a plan for attribution, notice preservation, data handling, and downstream access control
Do not begin commercial use until written permission is granted.
Legal note
This page summarizes FormosanBank's public position. It does not override stricter corpus-specific terms, source licenses, contributor agreements, community restrictions, or written agreements.
Last updated