Is there way to provide tip-offs relating to content on the database?

Support & Questions FAQ
1

Hi. Is there a way to provide anonymous tip-offs about mistakes, duplications, missed information regarding companies held in the database?

2

The answer to that is no. OpenSanctions is based exclusively on information that’s published by (trusted) third parties, usually governments. It’s a documentation of “known knowns”, not an original investigative product.

If you have confidential tips, I would propose sending them to a well-known investigative group in the relevant country. This is a great overview of the more hard-core outlets in each place: Our Members – Global Investigative Journalism Network

1 Like
3

I wanted to add a few comments here that may be helpful to others that find this thread. Pudo has confirmed above that OpenSanctions is not an original investigative product. For that reason, there is essentially no way to contribute data directly to OpenSanctions and that is unlikely to change. DStewart was asking about anonymous interactions but my comments below are about contributing in general without anonymity.

There is a public spreadsheet for reporting duplicates and in my experience the OpenSanctions team resolve these quickly - arguably that is one narrow exception where we can contribute directly to improving data.

It is possible to contribute indirectly by contributing to one of the sources OpenSanctions uses. The most obvious example would be correcting errors and/or adding content to Wikidata. In fact, OpenSanctions has a tool called Poliloom to encourage this. If you spot an error in a government database or other non-wiki source you can of course write to the publisher and ask them to fix it.

Those comfortable with code can also propose changes to source code on GitHub which could potentially help with de-duplication for example. See: OpenSanctions · GitHub. I think is right to acknowledge here that de-duplication is a complex problem with no easy answers. Some organisations for example have very similar names to each other and the risks of an incorrect match are often more serious than the risks of missing a valid match.

1 Like