"Ownership" of Data
- TYPE: Combined Session DATE: Thursday, May 17, 2018 TIME: 15:30-16:30 LOCATION: AMMERSEE II
Informational Self-Determination in a Post Facebook/Cambridge Analytica Era
Nat Sakimura from Nomura Research contributed his following viewpoint which will be discussed in this session:
The hype and hysteria around blockchain, blockchain identity and Facebook/Cambridge Analytica scandal have been quite interesting to watch. It did and is still showing a lot about people's understanding of the space, which is actually a bit different than what I think.
For example, people think that they own their data and ought to have sovereignty over their data and some people think that Blockchain identity and DID can achieve it.
I do not think so in general.
It is because of three reasons.
- The assertion that we own our data is wrong.
There are very few data that falls into "I have full right to control" category. Most of the time, the data is actually shared among people, as they form relationships. For example, your DNA sequence is shared with your relatives. You do not have right to disclose it publicly as the result. Your location data is shared with someone you are with, and disclosing your location would disclose her location as well. Do you have a full right to disclose it? Probably not. Not only because people do not read the terms, because of this shared data aspect, "consent" is an unreliable mechanism for the data processing. And the Facebook/Cambridge Analytica scandal's root cause actually is here. It was not a hack. People "consented" to provide "his data". His data, in this case, included data about his friends.
- I do not see anything particularly new in SSI
The basic model of SSI, as I understand, is that you write your identifier and claims location on a Blockchain. So, the blockchain works as the registry. People can then search the registry to find the location of the associated claims. Claims are not written to the blockchain. It is hosted off-chain. Does it not resemble something? It has just replaced DNS with a consortia run blockchain and Identity (=set of claims) Provider with Claims Provider. It just looks to me like the same model with new tools and some nice marketing phrases. Since it is easy for the authority to take down the claims provider, in such a situation, it is likely that the guy will be left only with his identifier, which is rather useless. Worse, the fact that most crypto-currency traders do not manage their keys themselves but use "online wallets" provided by cryptocurrency exchanges will make me think that they will probably use the Claim Provider as the online wallet and we are back to the square one. Welcome to the good old Online Identity Providers.
Compared to this, Self-issued IdP (SII) in OpenID Connect looks much more radical. We got rid of the registry. It is completely distributed. It lives on your handset. We do not need a shared database like blockchain to find claims providers because the SII can provide the claims or claims locations locally. These claims can be signed by the source so it is verifiable as well. It can be deployed without blockchain so we do not have to worry about the numerous technical issues of the blockchain that are not solved yet. Actually, Cardspace was on a similar model.
- There are no economic incentives for RPs and Users to start using it.
As I explained above, this "self-issued" model is not new. SII has been there since 2014 and I know of only one large-scale deployment (It started this February, by the way). Cardspace was even pushed through Windows 7 installations and it still did not fly.
Why? It probably is because there are no incentives for RPs to accept self-issued identities while the population coverage is not large. The investment to start accepting SII cannot be justified. The converse is true for the users. If there is no RP, then there are no incentives for the users to install and use SII. It is a classic chicken-and-egg problem.
Do you remember how Google got their identity flying? It was through a killer RP service called Gmail. After there were enough users, then RPs started to have incentives to start accepting Google identity. The same applies to Facebook.
Unless there is a way for the Self-sovereign identity to break through this problem, I do not see any reason why it should fly.
- Registration fee:
- Contact person:
Mr. Levent Kara
+49 211 23707710
- May 15 - 18, 2018 Munich, Germany