instangle commitment

Brand Safety Policy

instangle generates content angles at scale. That means we have a categorical responsibility to refuse any angle that would harm the user of our product or the audiences they reach. This policy is enforced at the model level, not as a post-filter. Stories in any of the categories below are refused before they ever enter your dashboard.

Categories we refuse by default

Tragedies involving named victims. No angle will be generated on any story involving identifiable individuals who have died, been harmed, or been victimized, regardless of how commercially relevant the broader story may be.
Active armed conflict. Wars, military operations, and regional violence are refused while the situation is unfolding. Retrospective analysis is permitted only after official resolution.
Partisan elections in flight. From the official campaign start through election certification, we do not generate angles on elections or named candidates in any country.
Contested medical claims. Claims about treatments, pandemics, or individual health outcomes that are not supported by current peer-reviewed consensus are refused.
Minors as subjects. No angle will target, identify, or reference a minor by name, school, or identifying characteristic.

How refusal works

When a story in any of the above categories breaks, it is flagged at ingestion and tagged with a refusal reason. The story never enters your niche-mapping queue. Your dashboard shows a count of refused stories for the day so you know the firewall is working. You cannot override the refusal. This is a hard rule, not a preference.

What you can appeal

If you believe instangle has refused a story in error (for example, a retrospective analysis flagged incorrectly as active conflict), you can submit an appeal through your dashboard. Appeals are reviewed by a human within one business day. Approved appeals unlock the story for your account only, not for the general index.

Why we publish this

Every AI tool that generates text about current events carries the risk of a single bad output destroying a user's credibility. We would rather refuse 5% of stories that might have been fine than generate one angle that connects a client's brand to a tragedy. This policy is the contract. If you run into an edge case, we want to hear about it. Write to safety@instangle.com.