Engineering contact-tracing apps has been a marathon
Why make them?
* manual contact-tracing became totally overwhelmed with covid cases
* can we supplement with technology?
Constraints: security and privacy
* protect from misuse: surveillance, target marginalized individuals, etc.
* purpose limitation by default
* hide user's identity, location, behaviour
* preserve system integrity
Those fears were well-founded -- information collected for covid was used by other parties e.g. police
Hidden constraint: reality
* highly scalable and reliable
* design under time pressure, has to be built and verified quickly
* dependencies, dependencies -- have to use the sensors we already have
First idea... never got made because reality is hard and has to work with the phones we have
* can't use too much battery
* Apple needed to be involved so that the app could see BT beacons (blocked for privacy reasons)
* needs to be compatible between Android and iOS
Google and Apple did the implementation, which had a lot of implications
* App creates a secret every day
* creates unlinkable BT tokens to broadcast that day
* when you see BT tokens, store them for later use
* If you're infected, upload your BT tokens to a server for when you were infectious
* Everyone gets the "infectious" tokens. If you've seen them, then the phone tells you to isolate and get tested
Strong privacy protections
* unlinkable
* information virtually all stays on the phone
* system sunsets by design
Protocol is just a part of a whole interlocking solution
* epidemiological etc
Authorization mechanism first design
* has to be hard to delegate
* privacy
We know how to do this with crypto (commit in a token)
... but health systems don't have the tools to handle this!
have to use a simple activation code
Are we done?
* in this system the only uploads are from covid+ people
* we can solve! first version: have everyone upload random crud
* but in the real world, you don't know peoples' behaviour. It's going to get messy. And you could accidentally DDoS the server/network
* battery
* also can't use an anonymity network (e.g. ToR) because it's not everywhere, would get DoSed, and we can't afford delays to notifcation (the whole point is to be faster than manual contact tracing)
How to get this access code?
Servers don't live in a vacuum
[ I can vouch for this. It's very messy. No you can't read that whole diagram, I know. It's messier than that. ]
Where is this deployed?
Field experiment in Zurich Oct 2020
* 80% of COVID+ users upload codes
* 22% sent quarentine
* 1/10 tested + after notification
Key lessons:
* data is not a must
* privacy engineering goes well beyond crypto
* privacy engineering in a cloud world is hard [yes]
* keeping in mind socio-technical integration is key
* purpose limitation and abuse prevention is a must
[end of talk]
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Last talk of #enigma2021 by Marcus Botacin: "DOES YOUR THREAT MODEL CONSIDER COUNTRY AND CULTURE? A CASE STUDY OF BRAZILIAN INTERNET BANKING SECURITY TO SHOW THAT IT SHOULD!"
The outcomes I get from my analysis of malware I find in Brazil were quite different than what I saw in analysis of malware from other researchers. Why? Because the malware attacks were different!
The Brazilian banking system:
* let's move banking to computers (80s)to keep up with hyperinflation
* desktop clients for users... and the attackers migrated from physical to fake desktop app attacks -- that would only work in Brazil because that's where the banking was
Content note: this talk is about online abuse as some of the content may be upsetting
Got pulled into this after a screenshot of a class assignment sending folks to post on 4chan to post about race/gender/etc issues got posted on 4chan without the email address... so the 4chan folks thought it was @gianluca_string. It wasn't, but they doxxed and harassed anyway
Kicking off the last session of #enigma2021, @katestarbird is speaking about an extremely pressing topic: "ONLINE RUMORS, MISINFORMATION AND DISINFORMATION: THE PERFECT STORM OF COVID-19 AND ELECTION2020"
This talk is going to go through the experience pushing the boundaries on sandboxing in the Chrome browser
What is sandboxing?
* breaking something into lower/higher privileged process
* necessary for browers, OSes, VMs etc.
Chromium uses to reduce the amount of privilege of the application: also to reduce the amount of privilege for code that touches websites (renderer)
* split different websites into different processes
* good defense against logic bugs (e.g. same-origin policy)
Next up at #enigma2021, Alex Gaynor from @LazyFishBarrel (satirical security company) will be talking about "QUANTIFYING MEMORY UNSAFETY AND REACTIONS TO IT"
Look for places where there are a lot of security issues being handled one-off rather than fixing the underlying issue
We tried to fix credential phishing mostly by telling people to be smarter, rather than fixing the root cause: people being able to use phished credential.
Zoom's launched end-to-end encryption 5 months after the white paper was published
* prevents eavesdroppers between users who are speaking to each other
* protection against compromised servers