william manning wrote on 2019-04-05 09:43:
Every now and then, Paul Vixie and I are in complete harmony.
i am in no way concerned about that.
In my current slot, we are one of thousands of entities that are being held accountable to a series of regulatory requirements that have significant fiscal impacts on the exfiltration of private/patient data. We are starting to focus on three distinct areas to reduce the impact that DOH presents to our security posture. ...
sadly, there are some here, and many elsewhere, who consider that you already had that burden, because the opacity of HTTPS especially with TLS 1.3 and encrypted SNI, means that the exfiltration risk preexisted, and was not made worse by DOH.
those considerations are naive and incorrect. however, it's necessary to explicitly re-dismiss them every time you mention the imposed costs of DOH. it is the _standardization_ aspect of DOH, and the possibility of encountering it inside HTTPS TCP IP DST addresses that did not offer it pre-standardization, that imposes the _new_ exfiltration and other risks.
This genie has not signed BAA or supplier agreement with us and we will not allow it to dictate our business processes or affect our liability without the DOH enabler shouldering fiscal and legal exposure when DOH is shown to be the culprit in exposure of private data. I can't see how DOH is going to pass GDRP muster inside the EU either, but that is for others to debate. I have told my GDRP affected counterparts about the privacy risks with DOH deployment.
i hear your pain. -- P Vixie _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
