[euopendata] Moral clauses in data terms & conditions?

Wed May 28 10:07:09 UTC 2014

> Ton Zijlstra <mailto:ton.zijlstra at gmail.com>
> 28 May 2014 10:17
> Hi Michael,
>
> Thanks!
>
> "Is it possible to gain insight on people because we have enough
> granular data sets?"
>
> This is a key issue, regardless of illegal or immoral use of data. The
> short answer to this is of course "yes". And none of our current
> dealing with privacy and data protection in laws takes it into
> account. Recombination of datasets that themselves are innocent into
> something that breach privacy does not exist as a possibility in
> current privacy law. Most deal only with whether a single dataset
> contains personal data.

 This is only partly true. The famous Google ruling on the right to be
forgotten was based on the emergent properties of the search results,
which amounted to a new privacy risk separate from the sum of its parts.
This made Google a data controller, while the original newspaper holding
the article was not affected.

It is a bit different when merging datasets, but I believe the principle
applies.
>
> We know that from your non-public FB profile that just lists 8 of your
> random contacts, a reliable (but not fully reliable) prediction about
> your sexual orientation can be made. Something any Ugandan male for
> instance should regard with fear.
>
> This is a question that will undoubtedly will be extremely important
> in the coming years in the (open) data discussion.
> Our notions of privacy and how we aim to regulate it will need to
> evolve a lot to keep up with the new affordances of pervasive digital
> data availability.
>
> best,
> Ton
>
> ---------------------------------------------------------------------------
> Interdependent Thoughts
> Ton Zijlstra
>
> ton at tonzijlstra.eu <mailto:ton at tonzijlstra.eu>
> +31-6-34489360
>
> http://zylstra.org/blog
>
> ---------------------------------------------------------------------------
>
>
>
> _______________________________________________
> euopendata mailing list
> euopendata at lists.okfn.org
> https://lists.okfn.org/mailman/listinfo/euopendata
> Unsubscribe: https://lists.okfn.org/mailman/options/euopendata
> Michael Hörz <mailto:hoerz at michael-hoerz.de>
> 28 May 2014 09:43
> Hi Ton,
>
> thanks a lot for your extensive input. That's very helpful.
>
> Briefly on your question: "And what would you want to teach in that
> regard?"
>
> It is partly ironic, admittedly. We're taking a somewhat playful
> approach on the parts of open data usually not covered. Is it possible
> to gain insight on people because we have enough granular data sets?
> How could "bad" be done with open data?
>
> Terms such as "bad" also involve moral assessments, therefore I'm
> interested how far such statements can be found in legal texts, where
> they are actually useless, as you pointed out.
>
> Best,
> Michael
>
>
> _______________________________________________
> euopendata mailing list
> euopendata at lists.okfn.org
> https://lists.okfn.org/mailman/listinfo/euopendata
> Unsubscribe: https://lists.okfn.org/mailman/options/euopendata
> Ton Zijlstra <mailto:ton.zijlstra at gmail.com>
> 28 May 2014 08:11
>
> Hi Michael,
>
> some data points:
>
>
> The national Norwegian early draft national license used to have a
> statement like that
> (http://www.epsiplatform.eu/content/norway-discusses-psi-licensing-model),
> but they moved away from it quickly
> (http://www.epsiplatform.eu/content/norwegian-data-license-feedback-request)
> and the final version does not have such a statement
> (http://data.norge.no/nlod/en/1.0).
>
> Dutch government bodies who think they need their own license
> sometimes add it, but the national policy is CC0 and/or Public Domain.
>
> The Flemish government (Belgium) has a suite of licenses (in which CC0
> is the preferred option), for gov bodies to choose from. Their 'by
> attribution'  and other licenses (paid for data, paid for commercial
> reuse, who are both discouraged by Flemish gov) however does state
> "The licensee commits to avoiding any unlawful re-use of the product"
> (http://www.opendataforum.info/files/Modellicenties_ENG.pdf)
>
> The French national open data license does not mention 'illegal' use
> but does state:
> "The «Re-user» is solely responsible for its re-use of the «
> Information »." http://wiki.data.gouv.fr/images/0/05/Open_Licence.pdf
> As it obviously should be.
>
> In general I suspect inclusion of the 'illegal use' clause is born out
> of the data holders fear for liability, not out of wanting to regulate
> reuse.
>
> In that light: The Dutch government's standardisation body issued a
> report on open data liability risks for government bodies (in Dutch
> only:
> http://www.forumstandaardisatie.nl/fileadmin/os/documenten/Rapportage_open_data_en_aansprakelijkheid_def.pdf)
> which rules out risks for gov bodies due to malintented reuse of data
> they publish placing that responsibility on the shoulders of the reuser.
>
> best,
> Ton
>
> ---------------------------------------------------------------------------
> Interdependent Thoughts
> Ton Zijlstra
>
> ton at tonzijlstra.eu <mailto:ton at tonzijlstra.eu>
> +31-6-34489360
>
> http://zylstra.org/blog
>
> ---------------------------------------------------------------------------
>
>
>
> _______________________________________________
> euopendata mailing list
> euopendata at lists.okfn.org
> https://lists.okfn.org/mailman/listinfo/euopendata
> Unsubscribe: https://lists.okfn.org/mailman/options/euopendata
> Michael Hörz <mailto:hoerz at michael-hoerz.de>
> 27 May 2014 22:44
> Hi all,
>
> I'm presently teaching computer science students in Berlin on the
> subject of open data.
>
> We will be looking at the "far side" of open data (i.e. using open
> data for not so good things).
>
> In this context I'm interested in examples of moral clauses in terms &
> conditions for using data.
>
> Such as "not use any of the Information for an illegal or immoral
> purpose"
> (http://dashboard.glasgow.gov.uk/ckansupport/open.glasgow.gov.uk_Licence_version%201.0.htm)
>
>
> So if you know of such terms, especially with interesting wording,
> please share them.
>
> Thanks a lot,
> Michael
> _______________________________________________
> euopendata mailing list
> euopendata at lists.okfn.org
> https://lists.okfn.org/mailman/listinfo/euopendata
> Unsubscribe: https://lists.okfn.org/mailman/options/euopendata
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/euopendata/attachments/20140528/34ede2bc/attachment-0003.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: postbox-contact.jpg
Type: image/jpeg
Size: 1308 bytes
Desc: not available
URL: <http://lists.okfn.org/pipermail/euopendata/attachments/20140528/34ede2bc/attachment-0006.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: postbox-contact.jpg
Type: image/jpeg
Size: 1098 bytes
Desc: not available
URL: <http://lists.okfn.org/pipermail/euopendata/attachments/20140528/34ede2bc/attachment-0007.jpg>