Methodology & Disclosures
Last updated: 2026-08-20
Everything published here is for informational purposes only. It is not investment, financial or legal advice. We are not brokers or licensed advisers, and we do not issue buy or sell recommendations, price targets, or entry and exit points.
Where the content comes from
We follow 31 financial news sources over RSS. Stories describing the same event are clustered into a single event, so it appears once with the number of sources that covered it rather than repeating.
What the AI does
It reads the story and produces an Arabic and English summary, a sector classification, a sentiment direction, an impact score, a confidence score, and the assets mentioned. Its instructions require observational language — watch, expect, may, could — and explicitly forbid buy, sell, short, long or any imperative investment command.
The AI gets things wrong. A summary can misread a story, and an impact score is an estimate, not a measurement.
The impact score
A 0–100 number computed from the model's impact estimate multiplied by its confidence, then adjusted by how many sources corroborated the event, the trust weight of the source, how old the story is, and — where available — the actual price move on the most-affected asset after publication. It is a reading-priority ranking, not a price prediction.
Halal screening
Compliance flags on assets are based on public criteria and third-party data. They are not a fatwa. For religious rulings, consult a qualified authority.
Commercial disclosure
- Sponsored content: any sponsored card or placement is clearly labelled. Advertisers do not influence the editorial analysis or the ranking of stories.
- Broker links: some broker links may earn us a commission on signup or deposit. This does not change what we show or how it is ranked, and is not an endorsement of the broker.
- Subscriptions: paid plans raise limits and unlock the daily brief. There is no "premium" recommendation withheld from free users and sold as a tip.
Corrections
If you see an analysis that is wrong or misleading, write to support@signalnews.app and we will correct or withdraw it.