Colleagues

I have listened to your comments in recent discussions and had some preliminary talks with the RIPE NCC about what could be implemented. So now we have a second version of my proposal on personal data.

I was getting comments from people that LEAs need addresses for their investigations, but also people had serious privacy concerns about publishing their home address in the database. I was considering the idea of publishing addresses with restricted access. However it seems that  it is almost impossible to confirm the identity of LEAs (even from within our own region, never mind globally) and perhaps other NGOs that could be eligible to access this restricted data. So that idea is not viable.

Then there was an issue over data quality of the addresses currently published in the database. In the ORGANISATION object the address is assumed to be a postal address. This mandatory address is entered and managed by the resource holders. No verification is done on this address. In the database documentation it says:
"This is a full postal address for the business contact represented by this organisation object."
That could be anyone. They don't even need to be located in the country where the organisation operates it's business. Being so loosely defined, any kind of verification would be impossible. That makes this address almost meaningless. It is on the same level as the "country:" attribute in resource objects. It only has meaning to whoever manages the data. Also by having this unverifiable address mandatory, we are almost inviting those who don't want to be easily located to enter false data. Especially as it is almost impossible to identify any of this postal address data as being true or false.

I am therefore suggesting we make this postal address an optional attribute. If any resource holder wishes to enter this optional postal address they can do so. But if it is a personal address they must not enter more than region and country. The full personal address of a natural person must not be entered into the database in any object type. Optional data, when provided, is more likely to be accurate. Making this address optional is in line with a recommendation of the RIPE Database Task Force.

Some will still argue that false data is useful if 'bad actors' enter the same false data in different places. That offers investigators an opportunity to cross reference this (false) data over different objects using the free text search facility of the database. In some cases this may be the only way to make these data links. Doing this has many problems. We cannot justify preserving false data in the database to allow some people to use an 'accidental feature'. There is no guarantee the same false data will be used in multiple places. There is a defined purpose of the database that allows LEAs to use public information from the database as part of their investigations. This purpose is actually about granting permission to LEAs to use available data. It does not define any data to be published in the database for the sole benefit of LEAs. So there is no purpose requiring this postal address to be published in the database. Optional is therefore a convenience for anyone who wishes to enter it.

cheers
denis
policy proposer