Content Policy

1. Why we screen

Oreni is a structured-learning platform. We accept content that fits a teach-and-learn format and that meets the platform’s safety and quality standards. We reject content that does not fit either of those criteria. Screening runs before generation, so rejected URLs never produce a course.

2. Accepted content

Oreni works best with — and accepts — videos that are:

• Teachable in format: tutorials, lectures, demos, structured talks, documentaries, and well-segmented long-form interviews.
• In English for the audio and on-screen text. (Other languages are on the roadmap.)
• Reasonably long: at least five minutes, and not longer than six hours. Videos over one hour additionally need chapter markers or timestamps for the generation pipeline to segment them reliably.
• Completed recordings: live streams and upcoming premieres are not yet supported.

3. Rejected content

The platform does not generate courses from videos that are:

• Not teachable: reactions, gameplay streams, vlogs, reviews framed as opinion content, drama or commentary channels, ad reels, or short-form entertainment.
• Adult or sexually explicit, including content that features sexually suggestive performance or imagery as a central theme.
• Centred on gambling: casino play, poker, sports betting, lottery promotion, betting-app walkthroughs.
• Centred on alcohol or recreational drugs: tasting, brewing, reviewing, or promoting alcohol; cannabis or other recreational-drug-centric content.
• Occult, divination, or pseudo-spiritual practice: tarot, astrology, palm reading, psychic claims, witchcraft, occult ritual, chakra and energy-work, third-eye and astral-projection content.
• Centred on music as performance: songs, music videos, vocal performances, instrument lessons, and DJ sets. Videos where music is incidental background to an unrelated topic (e.g. a coding tutorial with light background music) are not affected.
• Harmful pranks, mockery, or harassment: pranks that terrify or humiliate people, content that mocks individuals or religious communities, public-freakout compilations.
• Manifestly misleading or fraudulent: scams, pyramid schemes, fabricated news, manipulated medical claims.
• Cooking videos centred on pork or its derivatives (bacon, ham, prosciutto, lard, gelatin sourced from pork).

4. How screening works

Screening is automated and uses three layers, in order:

1. A fast rejection layer applied to video metadata (category, title, duration, live status).
2. A classification layer that inspects the title, description, and metadata of the video with an AI model and returns a structured decision.
3. A cache that remembers the decision for each video ID, so repeat submissions don’t re-run the AI step.

The user-facing rejection message is intentionally short — “This video isn’t supported on Oreni.” — and does not enumerate the category that triggered the decision.

5. Appeals and false-positive reports

Rejection decisions are not appealable in the sense that there is no override request you can submit to force generation on a rejected URL. However, if you believe a video has been misclassified you can report the false positive from any course page via the in-app “Report inaccuracy” button (select “other” and paste the rejected URL in the note). The team reviews false-positive reports weekly and adjusts the screener.

6. Acceptable use of the platform

Beyond the content categories above, you agree not to use Oreni to scrape, mirror, redistribute, or commercialise copyrighted material you do not have rights to. See the Terms of Service for the full list of acceptable-use restrictions and the Copyright / DMCA page for takedown procedure.

7. Changes to this policy

The content policy is a living document. We update categories as we observe edge cases and as the screener matures. Material changes are reflected in the “Last updated” date at the top of this page.

8. Contact

Questions about a specific rejection or about the policy itself: email contact@oreni.co with the URL and a short note.