I’m at #AMCTO2019 in Huntsville this morning presenting findings from my upcoming report on the use of online voting in Ontario cities. We may be the largest global user of unregulated online voting, and that needs to change! @AMCTO_Policy@ElectionsON
Online voting went offline for a million voters on election night due to a miscommunication, lack of disaster plans, and no apparent load testing. We must do better for 2022.
We found the use of dates of birth as login credentials uniquely identified up to 1.3 million voters in their respective towns. This wouldn’t be acceptable in the health data world, and should not be acceptable in a secret ballot election.
As voters become increasingly aware of technology, they want to know how it’s protecting their ballot, and candidates deserve objective evidence of the outcome. Standards development is desperately needed!
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Setting aside the dubious normalization of ubiquitous ID checking, vaccine passports create new opportunities for inescapable data collection 🧵
Proponents of vaccine passports rightly point out that showing ID was something we previously had to do, like when buying alcohol or entering a bar
However, in those settings, a human looks at your ID. They don't record it. The interaction is ephemeral. They make a decision in the moment, then it's gone
I don't use remote online proctoring services in my courses. As a cybersecurity professor, I couldn't in good conscience make my students download and install something on their device that I wouldn't install myself. lfpress.com/news/local-new…
I couldn't in good conscience require my students to submit to facial recognition software when I wouldn't myself. Or grant such an app system-level privileges.
I hear the term "we're confident" thrown around altogether too much in the context of someone else's data and someone else's device. What I never hear is the indemnification. I never hear the pledge. Don't tell us how you're confident. Tell us how you're liable.
Compare: tiny ovals, excessive white space, implicit oval/candidate associations
Here's a thread debating whether this mark would be counted by a machine, which, you'll notice, is a very different question than "should this mark be counted?"
1/5 Online voting vendor @Voatz has been engaged in an alarming campaign in essence to become the gatekeepers of their own cyber accountability. Today I join 70 security experts in a letter admonishing their recent submission to the @USSupremeCourtdisclose.io/voatz-response…
2/5 Our response has an unusually diverse list of signatories from academia, industry, and government. It includes those who work in software security in general, as well as those who work in elections in particular, making the case that @Voatz's views are not widely held
3/5 In their amicus brief, @Voatz referenced their work with @Hacker0x01 as evidence of the "success" of their bug bounty program. Notable among the signatories of this letter, therefore, is @Hacker0x01 and a number of other bug bounty organizations
Remember when online voting vendor @Voatz referred a @UMich student to the authorities? Well now they're arguing to the @USSupremeCourt that the Computer Fraud and Abuse Act should not be narrowed to protect independent "unauthorized" security research supremecourt.gov/DocketPDF/19/1…
At issue is the question of whether independent cybersecurity research is necessary. @Voatz argues research and testing "can be performed by authorized parties" and that "unauthorized research" and dissemination of "theoretical security vulnerabilities" is "harmful".
.@Voatz describes bug bounties as "highly effective" and even references their past association with @Hacker0x01 while brazenly omitting that they were removed from the program for not "acting in good faith towards the security researcher community." cointelegraph.com/news/voatz-bug…
So @nicolejgoodman and I testified to @HoCCommittees#PROC that remote voting was doable for NON-SECRET votes. The report completely omits this crucial point and instead inexplicably recommends "conducting votes via SECRET ballots electronically"
We detailed in an @IRPP oped why NON-SECRET voting was necessary for verifiability. We submitted this breif to the committee and summarized it in our testimony. Incredibly, these arguments were all omitted from the report and our brief wasn't even cited. policyoptions.irpp.org/magazines/marc…
By selectively excluding key portions of expert testimony from the report that would have run contrary to the committee's eventual recommendation, I find myself sharing @CPC_HQ's assessment that the government "seemed committed to a specific outcome."