Thread by @LeaKissner

Next up at #PEPR21: Cryptographic Privacy-Enhancing Technologies in the Post-Schrems II Era

from Sunny Seon Kang, Data Privacy Attorney

Going to provide context on CJEU case C-311/18, aka "Shrems II"

This launched companies into a whole tizzy because they said that folks needed "supplementary measures". What the heck is that?

Without Privacy Shield, you can't transfer data from the EU to the US (thanks, Shrems I), because the US isn't considered to have "adequacy" [essentially strong enough protections under the law. People were pissed about Snowden]

So instead we (often) use Standard Contractural Clauses. They're valid, but not always sufficient:
* SCCs don't bind the governments -- they're not a part of the contract
* So post-Shrems II may need "supplemental measures" which can be technical, contractual, or organizational

So you can't just use SCCs or BCRs (Binding Corporate Rules). You may need technical measures as well. On top of all the other ones you should have.

But at first no one knew what they were. There's guidance from EDPB on 11/2020

So what's an example of a technical supplementary measure? Secure multi-party computation (MPC). [cf en.wikipedia.org/wiki/Secure_mu…]

This allows parties to do computation on encrypted data.

For example, we could use secret sharing to choose several random numbers that add up to your salary. The other people do the same. Then you compute the averages and aggregate. You get the average without seeing the input.

[There are other forms, for example garbled circuits]

[Note that this is only appropriate for certain kinds of computation. It's *very* expensive speaking as someone who has spent a bunch of time trying to use this in various applications at scale. Plus works *very* poorly if you have abusive parties who are trying to mess you up]

But hey, the guidance points out that this is awesome and suggests you use it.

[And it is awesome! Encrypted computation should totally be used wherever you can. We're getting better at it over time with cryptographic advances, plus increased computing power helps a ton.]

So you might think, what's the holdup? Why doesn't everyone use it?
[Nope! But most people don't have a PhD literally in encrypted computation. 😆 We should talk about this a ton.]

* Feat of uncertainty. Regulatory guidelines are tech-neutral and don't say "this particular tech meets the bar"
* Overhead and scalability. Cryptographic privacy tech may require "significant configuration" with the legacy system

[I'd say this understates it. I tried to use encrypted computation for something once and Google literally didn't have enough computers available to do it. We would have had to shut down something like Gmail. This is *not* a small number of computers and this was one feature.]

Q: Who will be involved in making those technical guidelines?
A: Regulators and stakeholders and people who actually make these technologies. Also consumers are impacted and they may have feedback. Many multi-stakeholder meetings.

Q: What are the benefits/drawbacks of using 2 or more independent parties?
A: Wouldn't say there are more benefits -- use as many as you have participants.

[ This is not 100% true. You can in some cases use more parties to increase robustness against collusion. Or warrants. But be aware that this usually comes with greater network and computation cost, and more risk that some of your parties will have network or other failure]

My hands tell me it's time to take a break from livetweeting, so I'll see you at the next #PEPR21 session.