Fair Use & Attribution

Effective Date: May 7, 2026 | Last Updated: May 7, 2026

Corvus Intel is a real-time news and intelligence aggregator. It fetches publicly published RSS and Atom feeds from 800+ sources, normalizes them, and re-presents them in a dashboard with optional AI summarization and watchlist matching. This page explains the legal basis on which we do that.

RSS / Atom aggregation

RSS and Atom are open syndication standards. Publishers who expose a feed do so to invite third-party clients (newsreaders, aggregators, search indexers) to fetch and re-present headline-and-summary content per the feed's published terms. Corvus Intel:

• fetches feeds from URLs the publisher has chosen to make public;
• respects each feed's ttl / updated / Cache-Control hints;
• honors robots.txt on the source domain when available;
• presents only the title, link, summary, and metadata fields the publisher chose to include in the feed;
• links every item back to the original source URL so readers visit the publisher.

If a publisher does not want to be included, they can either remove the feed, change its terms, or contact us to be removed from the source list (see DMCA contact).

U.S. fair use (17 U.S.C. § 107)

Where AI summaries, briefings, or analysis quote portions of source articles beyond what the publisher's feed itself contains, that quotation is fair use under 17 U.S.C. § 107. The four statutory factors:

1. Purpose and character. Use is transformative — the AI synthesizes dispersed reporting into a structured analytical brief, watchlist match, or threat-trigger evaluation, with novel commentary and cross-source context. Purpose is news reporting, commentary, and research.
2. Nature of the work. Source material is overwhelmingly factual reporting on matters of public concern (geopolitics, defense, markets, energy, public health, technology, regulation). Factual works receive thinner copyright protection than fictional works.
3. Amount and substantiality. Excerpts are headline-and-figure-level. The user's own AI provider (their API key, their account) processes the corpus and returns summaries; we do not retain or redistribute the original article body.
4. Effect on the market. Briefings and watchlist matches send traffic upstream to publishers via outbound links. They are a substitute for nothing — the original article remains the canonical, full-detail source.

For non-U.S. readers, equivalent doctrines apply: UK fair dealing for criticism, review, and reporting current events (CDPA s. 30), EU quotation exception (InfoSoc Directive Art. 5(3)(d)), Canadian fair dealing (Copyright Act s. 29).

Nominative use of trademarks

Names of news organizations, government agencies, defense contractors, market participants, central banks, AI providers, and historical figures referenced anywhere in the dashboard are used under nominative fair use: only to the extent necessary to identify the underlying actor or source, with no implication of sponsorship, endorsement, or affiliation. The classic three-factor test (New Kids on the Block v. News America Publishing, 1992) is satisfied — the entity is not readily identifiable without using its name; only as much of the mark is used as needed; and nothing in the dashboard suggests endorsement.

How we attribute

• Every feed item displays the original publisher name and links to the original article URL.
• Every AI summary, watchlist match, briefing, or threat assessment is generated against a corpus of attributed feed items pulled from the public-source list.
• The default source list is documented in the dashboard and can be exported as OPML.
• When AI features are used, requests go directly from your browser to the AI provider you have configured (Anthropic, OpenAI, Google, Groq, Perplexity, or your local Ollama). The provider's terms of service and data-handling policies govern that connection.

What we do not do

• We do not republish full articles.
• We do not paywall-bypass any publisher.
• We do not strip bylines, datelines, or publisher attribution from feed items.
• We do not host article body text on our servers (cached content lives on your device in IndexedDB).
• We do not sell, syndicate, or relicense third-party reporting.

Citing Corvus Intel

Suggested citation:

Corvus Intel. (2026). [Page or briefing title]. Retrieved [Date], from https://corvusintel.app/[path]

If your work uses our default-source list or our briefing methodology, please link back so your readers can audit the underlying corpus.

DMCA / takedown contact

If you are a copyright owner and believe content displayed via Corvus Intel infringes your copyright, send a notice meeting the requirements of 17 U.S.C. § 512(c)(3) to:

Include: identification of the work, identification of the allegedly infringing material with URL, your contact information, a good-faith statement, a statement under penalty of perjury that you are authorized to act, and your physical or electronic signature. We process valid notices within 10 business days. If a feed source asks to be removed from our default source list, we will remove it within the same window.

Corrections

Spotted an error? Contact us with the page URL, the source URL, and a brief description.

Home · About · Team · Privacy · Terms · Cookies · Disclaimer · Accessibility · Contact