5 ms·
The main parts of the wrong-doing: - The school had obtained consent from the pupills and each one of the pupills could opt out at any time. The Swedish Data P
by havkom 7y ago
The main parts of the wrong-doing:
- The school had obtained consent from the pupills and each one of the pupills could opt out at any time. The Swedish Data Protection Agency (DPA) did however find that due to the power balance of a school vs a pupill, these consents are not ”freely given”, and thus are void. The school therefor have no legal basis of processing special categories of data/sensitive data, which processing face recognition data (”biometric information”) for identification purposes is. Violation of GDPR Article 9.
- The school had not completed a sufficient Data Protection Impact Assesment which properly identified risks of this processing. In addition, due to the nature of this processing, they would have been obliged to consult with the DPA about this impact assessment before starting. Violation of GDPR Article 35&36.
- The DPA also found that the processing was more extensive than necessary, since attendance could be taken in less invasive ways. This a violation of GDPR article 5.1c (the data minimisation principle).
I hope the decision will be appealed by the school, which is run by the municipality. There are several interesting questions, such as if consent can be freely given in schools, if the data minimization principle was really breached (the school claimed this way of taking attendance saves a calculated 72,000h of teachers time per year in just this school), etc. I am not sure however that the municipality has the right resources and competency to make a competent appeal.
- Iv 7y ago> the school claimed this way of taking attendance saves a calculated 72,000h of teachers time per year in just this school Which is 197h per day. I find this number extremely dubious.
- nmstoker 7y agoThe figure I saw (on the BBC article and the linked Swedish article) was 17,000 hours per year in this school. That still seems suspect: estimating a school year of 35 weeks that's 175 days, thus 97 hours per day across all teachers. Even if it were some mega school of 1,000 pupils and maybe 100 teachers that seems high, would imply they spent perhaps ten minutes of every hour in a six hours day taking & dealing with attendance on average! (obviously based on assumptions, so this could vary) It feels like a number that was come up with (or at least inflated) to justify the project.
- TazeTSchnitzel 7y agoPerhaps it is a number based on the time that would be required for teachers to perfectly track attendance. But of course they don't in practice!
- havkom 7y agoMy fault. 17 280 h per year was the calculated savings
- SiempreViernes 7y agoThe new system could well have automated more than just the role call, and so they can add in savings in the post processing of the attendance records.
- labawi 7y agoPerhaps .. But what processing was enabled that could not have been done without camera surveillance?
- SiempreViernes 7y agoMy point is that the gain in time could we be mostly due to automating other parts of the system that are currently done manually.
- havkom 7y agoMy fault. 17 280 h per year was the calculated savings