OpenStreetMap
Volunteered Geographic Information … why, that's us!
Posted by scruss on 24 February 2016 in English.The Canadian Internet Policy & Public Interest Clinic (CIPPIC) has an interesting piece on the legal risks of contributing to and using Volunteered Geographic Information … in other words, OSM: Volunteered Geographic Information FAQ.
The document makes a few observations (from a Canadian legal perspective) that mappers might find useful. In summary:
-
As a mapper, you’re unlikely to be sued if you add incorrect data to OSM. But you wouldn’t (knowingly) do that, would you?
-
OSM (or OSMF, most likely) would unlikely be liable if map data were relied upon in an emergency and found to be incorrect. There are many nuances to this, so it’s better if you read the When would a VGI website be liable in negligence for inaccurate information? section yourself.
-
In order to limit its liability, OSM should “… also make sure that their software is running smoothly, and that the visual designs of the sites are clear and informative.” Hmm.
(Incidentally, CIPPIC is also responsible for the excellent-though-depressing CIPPIC Licensing Information Project for Open Licences, CLIPol. If you want to see how not to do open data, take a look at some of the really bad licences they’ve found. I wonder why so many of the terrible “open” government licences are near me?)
Discussion
Comment from SimonPoole on 24 February 2016 at 16:56
Hmm the underlying issue in the analysis is that it is somehow based on the assumption a) VGI has specific quality issues that “professional” information doesn’t have, and b) that the liability issues are significantly different for VGI based sites. Neither would seem to be true.
Comment from Sanderd17 on 24 February 2016 at 19:20
OSM is also protected by intellectual rights, just like commercial data is. The only difference is that we use intellectual rights to force a rather free license on the users of the data.
But one of those terms in the license (which anyone is required to agree upon when using the data), is that OSM isn’t liable for mistakes in the data.
See http://opendatacommons.org/licenses/odbl/1.0/ (point 7.1).
So I see absolutely no difference with other providers.
Comment from Sanderd17 on 24 February 2016 at 19:22
(and also section 8)
Comment from Warin61 on 25 February 2016 at 01:53
Hot Coffee .. with warning “Contents Maybe Hot” … ?
I accept no legal responsibility for the acts of Darwin!
Map makers have long included phantom objects (including link roads) to detect copyright infringements. A reasonable person would accept some errors and inaccuracy in any map or other document, to do otherwise is foolish. I suppose that is why ‘we’ have foolish laws.
Comment from scruss on 25 February 2016 at 15:14
To various points:
Yes, the assumption that volunteer data is of lesser quality is annoying. I should reach out to CIPPIC and point them to research like Anita Graser’s OSM quality assessment with QGIS: positional accuracy. Got any other links?
In a Canadian court, if the three tests described in When would a VGI website be liable in negligence for inaccurate information? were met, then the mapper could be found liable. If in the unlikely case of a link between that mapper and OSM(F) being found, then they too could be liable. The article makes the point that a licence disclaimer alone is not sufficient.
Canadian municipalities must publish/maintain up-to-date maps, by law. Any phantom details in OSM knowingly supplied or left uncorrected and used by municipalities could open them to liability. ‘Could’ is the operative term here; there are no precedents, as the article says.
Regarding the Darwin comment, almost everything you think you know about the Stella Liebeck case is wrong.