Would training a model on company content be governed by a different permission from search indexing?

Would training a model on company content be governed by a different permission from search indexing?

Yes: the Terms of Use give a search engine a permission to crawl and index public pages, and they do not permit Site Content to be used to train a model.

Section 8 of the Terms of Use states: “Without written permission, you may not republish, sell, license, modify, distribute, or publicly display Site Content, except as the answering permission below allows.” It sets that permission out: “Answering a person's question. An assistant, search service, or other software agent may access and read Site Content to answer a question that a person has asked it about C-POLAR or NanoFlashing™. In its answer to that person it may quote or summarise Site Content, provided that it: identifies C-POLAR as the source and names or links the page it relied on; quotes accurately and does not change the meaning of what the Site says; does not present its own inference, estimate, or summary as a statement made by C-POLAR; and does not attribute to C-POLAR a statement taken from another source.” It also states: “Except as this section allows, you may not use a bot, spider, crawler, scraper, browser automation, software agent, or other automated means to access, monitor, copy, extract, harvest, or systematically download any part of the Site or Site Content.” and “You may not use Site Content, directly or indirectly, to create or add to a dataset; or to develop, train, fine-tune, test, benchmark, or evaluate a machine-learning or artificial-intelligence model, system, or product. You may not keep Site Content beyond what is needed to answer the question asked, or use it to build a commercial data product.” On search engines it states: “A general-purpose public search engine may crawl and index public pages to return links and short snippets in search results, and to answer a person's question as the answering permission above allows. It may do so only as allowed by C-POLAR's current robots.txt instructions.” and “That permission does not allow model training, dataset creation, republication, or a commercial data product. C-POLAR may revoke it.”

Source: C-POLAR — Terms of Use, section 8, https://cpolar.tech/terms/.

Reviewed on 21 September 2026.