# EventIndex.ai — Data & takedown policy

Events are almost entirely factual data: what, when, where, how much. The policy
follows from that — we store the facts, we publish where they came from, and we
remove what should not be here.

## What we store

Where the name of a natural person appears (an artist, a promoter, an organizer)
we store the minimum: the public name and a link to the public source it came
from.

**Stored.** Title, times and timezone, venue and coordinates, price, category,
organizer name, public source URLs, verification timestamps.

**Never collected.** Scraped contact details. Emails and phone numbers do not
enter the pipeline at all — there is no field for them.

## Removal requests — 72 hours

If you are an organizer, or a person named in an event, and you want data
corrected or removed: write to privacy@eventindex.ai with the event URL.
Requests are processed within 72 hours. No form, no account, no ticket number.

| When | What happens |
|---|---|
| `T+0h` | Request received, event located by URL or `evt_` id. |
| `T+72h` | Record corrected or removed, and the source that supplied it is flagged so it does not return on the next sync. |
| after | The URL returns a plain 404 or the corrected record. We do not leave a stub that says something was deleted. |

## Provenance

Every event lists its public sources. Summaries are written by us and never
copied from a source. Facts are verified against sources on a rolling schedule,
and every page publishes `last_verified` — relative and absolute, in UTC.

Images are not republished. Version one of the index carries no event imagery at
all except first-party submissions, which is a copyright decision as much as a
performance one.

## Visitors

The directory sets no cookies and loads no third-party scripts, so there is no
consent banner to dismiss. The one preference we keep — whether you chose the
dark theme — lives in your own browser's `localStorage` and is never sent to us.
Server logs record requests, including crawler user-agents, and are used to see
which cities AI crawlers actually read. Nothing follows you between pages.

## Sources that ask to be excluded

A site that does not want to be read tells us so, and we stop reading it.
Ingestion sources are database rows: disabling one takes effect on the next
cycle, without a deploy. We ask the same courtesy we extend to AI crawlers — say
it in writing, and we will honour it.

---

HTML version: [https://staging.eventindex.ai/data-policy](https://staging.eventindex.ai/data-policy)