,
13 tweets,
6 min read
Read on Twitter
[Thread] How can we (continue to) align the internet infrastructure with the public interest?
I had the priviledge of discussing this question with bright minds @Harvard - many of whom I have been admiring for a long time (names later).
We did not come to a conclusion. But ...
I had the priviledge of discussing this question with bright minds @Harvard - many of whom I have been admiring for a long time (names later).
We did not come to a conclusion. But ...
... let me share some of the issues that were flagged and possible avenues for action that were identified.
[I am using passive voice because the workshop was held under Chatham House Rule en.wikipedia.org/wiki/Chatham_H….]
[I am using passive voice because the workshop was held under Chatham House Rule en.wikipedia.org/wiki/Chatham_H….]
Background: System designs translate social choices into technology. In the case of the internet, the way the technical protocols are designed can even condition how internet users can exercise their Human Rights (HR). So why don't we just hardcode HR into the internet?
Because it is complicated. There are two main reasons.
#1 is that "condition" does not mean "determine". Technical standards may condition action by suggesting certain outcomes, but whether these imaginaries become reality will also depend on how implementers use the standards.
#1 is that "condition" does not mean "determine". Technical standards may condition action by suggesting certain outcomes, but whether these imaginaries become reality will also depend on how implementers use the standards.
Take the example of DNS. The Domain Name System protocol (elders may think of it as the phone book service of the internet) is currently the only non-encrypted web protocol that everybody uses. Such openness makes it prone to eavesdropping, revealing which websites you visit.
If the few browser providers send all DNS requests to a handful of resolvers (i.e. the phone books), the operators of those resolvers amass gatekeeping power that could be abused for monitoring or censorship. This is detrimental to the intention of standardising for Human Rights.
A conclusion from this is that a) standardisation bodies must acknowledge the Human Rights dimensions of their work - protocols are not neutral. But b) we must put implementation on the table. Working with implementers may sometimes be more feasible to remove user's pain points.
There is another reason (#2) why standardising for Human Rights is less straightforward than it might seem. This is that in reality, Human Rights can come into conflict with each other. A polarising example is the issue of anonymity vs. real-name policy.
Anonymity is important e.g. for LGBTQI protection, while a real-name policy will help law enforcement track down violators. - Such cases are not necessarily contradictory in a political system where you can tie rights to roles.
Unlike in political systems, tying Human Rights principles to technical standards does not allow for compromise. It was argued that by definition attempts to encode HR into internet protocols locks trade-offs between HR that should be open to contestation.
E.g., one participant pointed out that current standardisation attempts such as encryption seem to favour freedom of expression over children's or religious rights, which could invoke HR protection as well. - As I said: It's complicated.
I will continue this thread later. Meanwhile, follow @beatricemartini, @nielstenoever, @C___CS, S Braman, @biellacoleman, @malloryknodel, @mehwishaansari, Daniel K Gillmor, @niftyc, @shivan_kaul, @joebeone, @teirdes, @btayeg, @isik5, @brennan_mike, @pranesh, @dourish, @tarakiyee
