9 ms·
Microsoft has processes in place to minimize business risk due to privacy and security problems (both usually go hand in hand). All engineers must go through re
by cmarschner 9y ago
Microsoft has processes in place to minimize business risk due to privacy and security problems (both usually go hand in hand). All engineers must go through regular training, and features must go through security and privacy reviews. In these reviews, systems and their threat boundaries are described as well as the data that flows over system boundaries or that are stored. PII (personally identifiable information) gets anonymized unless absolutely necessary (e.g. for a profile page). The mantra is that the engineers should trust the system to put their own data in. Then there is also GDPR which requires adding abilities to delete personal data on request, even if it is distributed into different systems. It‘s even tougher when it‘s about customer data (eg Office 365). That‘s an absolute black box, so to do ML on this data you might put training into a firewalled cloud and you might get a metric back, but there is no way of looking at any concrete data. So, MS treats the prospect of privacy breaches as an existential risk and acts accordingly. It has been this way for at least 10 years.
- pavlov 9y agoThe "lost years" spent on fixing Windows security fundamentals between XP and Vista perhaps taught Microsoft something important about long-term risk management.
- fauigerzigerk 9y agoIf it's so existentially important to them, why is OneDrive data in personal (paid) Office 365 accounts not encrypted at rest?
- Const-me 9y agoWhy would you want them to? Their web servers needs decrypted data in order to handle these office documents, so client side encryption is not an option. Server side encryption, where the server knows the keys, serves very little purpose because the threats it protects against are extremely unlikely. What it’s protecting against, armed robbery of the datacenter?
- fauigerzigerk 9y agoI want to be protected against the exact same attack vectors that Microsoft's business customers are protected against and the same attack vectors that all customers of Google, Apple or Dropbox are protected against. Perhaps it's all just security theatre as you suggest, but I am not the one running these data centers, so if all these companies say there is a threat, then my instinct is to believe them.
- Const-me 9y ago> I want to be protected against the exact same attack vectors that Microsoft's business customers are protected against I don’t think there’re such attack vectors. The reasons MS doing that with the data of their business customers are legal, not technical. They keep medical data, legal data, government data for many jurisdictions around the world. Before doing that, they had to comply with FIPS 140-2, HIPAA, and a whole bunch of others: https://products.office.com/en-us/business/office-365-trust-center-compliance-certifications https://products.office.com/en-us/business/office-365-trust-... > if all these companies say there is a threat I can’t remember them saying there is a threat, or mentioning what it is. I only heard them saying marketing BS about how safe my data will be on their servers.
- fauigerzigerk 9y ago>I can’t remember them saying there is a threat I doubt that any of them will ever publish their most detailed threat models. If I had to guess, I would say the most likely threat is nosy or corrupt staff. Protecting a key server is easier than protecting tens of thousands of servers and physical disks. Legal requirements are not necessarily baseless either. They are in place to protect someone against something. The data I have on these cloud services is extremely sensitive. All my identity documents, proof of address, examples of my signature, financial data, health data, etc. If someone were to get hold of these documents, they could steal all my money, my identity and make life hell for me.