Useful research has to survive inspection.

This policy applies to public educational pages, research notes and market analysis published under the Basis name. Product interface labels and automatically generated private user output follow the same accuracy principles, but may not receive editorial review before display.

Market claims should identify the relevant venue, instrument, timeframe and observation window whenever those details affect meaning. News claims should link to or name the original publisher. Secondary summaries should not be presented as primary evidence. Timestamps matter because crypto markets and derivatives conditions can change within minutes.

Basis uses AI to accelerate research, comparison and drafting. AI assistance does not make a statement true. Public analysis intended for publication must be reviewed against its cited data, checked for unsupported certainty and edited so that scenarios, assumptions and invalidation are clear. Private agent output remains user-directed research and should be independently verified.

We do not describe a signal as guaranteed, risk-free or certain. Backtests and examples must not be framed as expected future returns. Directional views should include the conditions that would weaken or invalidate them. Educational content is not individualized financial advice.

Sponsored placement, affiliate relationships or paid distribution must be disclosed near the relevant content. Editorial conclusions should not be sold to token issuers or venues. Provider attribution does not imply commercial endorsement.

Material factual errors should be corrected promptly. If an update changes the conclusion rather than only wording, the revised publication should identify that it was updated and why. Methodology and data-source changes are recorded by updating the review date on the relevant standards page.