Janoshik.com - Customer Data Security Incident

[janoshik]

So you kept the address the sample was shipped from for 2+ years? That's not great. But thanks for being transparent about the incident.

Also, since the data is probably out there on the dark web already. Any chance you could make the data available for academic research? Properly anonymized of course, with data use agreement, etc.

The customs and couriers keep the data forever.
Thus we deemed such data low risk.

The data by themselves are not very useful for scientific research - the samples are seldom random, the context is missing, most of the samples being blind the glaring issues are undetected without what the client knows, but we don't.

I do actually try to make the data useful for harm reduction, but other than some general statements I've had over the years in the interviews, there's not much to learn as far as I can tell.

 

[Walter Hype]

glad you told us

just a fyi guys always pay with untraceable bitcoin and a drop addy including buying your shipping labels with bitcoin or coin thats harder to track you will be good. if any of you need advice on how to purchase shipping labels using coins that is harder to track back to you ill assist you.

oh yeah and please use a name thats not you.......obvious but just a fyi

 

[amazonas]

glad you told us

just a fyi guys always pay with untraceable bitcoin and a drop addy including buying your shipping labels with bitcoin or coin thats harder to track you will be good. if any of you need advice on how to purchase shipping labels using coins that is harder to track back to you ill assist you.

oh yeah and please use a name thats not you.......obvious but just a fyi

Do you mean paying for the shipping labels with no KYC crypto cards? or prepaid VISA cards bought with crypto?

 

[MyNameIsJeff]

The customs and couriers keep the data forever.
Thus we deemed such data low risk.

The data by themselves are not very useful for scientific research - the samples are seldom random, the context is missing, most of the samples being blind the glaring issues are undetected without what the client knows, but we don't.

I do actually try to make the data useful for harm reduction, but other than some general statements I've had over the years in the interviews, there's not much to learn as far as I can tell.

I can think of all kinds of interesting descriptive stats you could extract. Graphs of the times series of purity results, for different types of PEDs, or for different countries. One could look for structural breaks in the time series around events such as government crackdowns. You could even try to run some regressions to try and explain the differences in purity results and/or the number of samples sent in from different countries, like the legal status there (dummy), population, density of gyms, geographical distance (shipping cost/customs related), etc., with time fixed effects if there's enough years. Oh yeah, one could also judge source popularity over time by the number of sent-in samples, control for their reimbursement policy, relate it to the probability of source exit (scam vs bust), which could be a nice addition to a game theoretic model. And there's more clever stuff which I don't want to get into here.

As I said, I'd be happy to agree to give you a right to review manuscripts prior to submission/publication to ensure confidentiality. You could also be a coauthor of course.

 

[janoshik]

I can think of all kinds of interesting descriptive stats you could extract. Graphs of the times series of purity results, for different types of PEDs, or for different countries. One could look for structural breaks in the time series around events such as government crackdowns. You could even try to run some regressions to try and explain the differences in purity results and/or the number of samples sent in from different countries, like the legal status there (dummy), population, density of gyms, geographical distance (shipping cost/customs related), etc., with time fixed effects if there's enough years. Oh yeah, one could also judge source popularity over time by the number of sent-in samples, control for their reimbursement policy, relate it to the probability of source exit (scam vs bust), which could be a nice addition to a game theoretic model. And there's more clever stuff which I don't want to get into here.

As I said, I'd be happy to agree to give you a right to review manuscripts prior to submission/publication to ensure confidentiality. You could also be a coauthor of course.

Well, we might work something out with the anonymized data.