It's like comparing Android to Qubes OS. Not really fair. For what they are, Google products are surprisingly secure.
For one their products are not "secured" from Google seeing your private data...
In simple terms, end-to-end encryption means that messages are encrypted on the sender’s device (before it even leaves their computer or mobile phone), and can only be decrypted by the recipient on their device. This means that no third party which transmits or intercepts the email between the sender and recipient (i.e. internet service providers, the NSA, or even ProtonMail as the mail server operator) can decrypt and view the message.
This powerful protection is possible because ProtonMail has PGP email encryption built-in. End-to-end encryption is done automatically without user interaction whenever messages are exchanged between ProtonMail users. For an enterprise using ProtonMail for their email hosting, this means all communications between employees are automatically protected with end-to-end encryption. ProtonMail can also support sending/receiving end-to-end encrypted messages with recipients who are not using ProtonMail. The use of end-to-end encryption makes ProtonMail a better choice for security conscious individuals and organizations.
Which would correspond to a tiny fraction of all email a normal user exchanges and Even corporate users would be unprotected if they were to use protonmail to communicate with third parties such as service providers.
It's not a panacea. There are physical threats, there are threats from the very hardware you're using. But, like it says on the box, it is a reasonably secure operating system.
[1] https://blog.quarkslab.com/xen-exploitation-part-2-xsa-148-f...
Also, with their recent foray into enterprise support, they will hopefully be able to expand their auditing efforts in the next couple of years.
That said, even though the argument is a bit flawed here, I think most attracted by it would still prefer ProtonMail for other sound arguments.
I don't know about any security brochure, but I know I can setup 2FA to use push notifications (not an insecure SMS number), and can check where all of my logins are from, and have "suspicious" logins blocked automatically, etc, etc. I can also create single-use passwords for insecure devices (such as a youtube password just for my apple TV). Not to mention how amazing they are about spam detection.
I think their security posture is actually excellent.
Yes. Most definitely. I don't use Gmail and I recently had a discussion with my coworkers about Gmail. When I asked why they use it one person said, "because it's Google" and another said, "What else would I use? Yahoo or Hotmail? Hahahah". A third person responded by saying , "it's just easy because it's Google so it connects to everything else from them."
Absolutely. No doubt in my mind.
What are some specific threats that Gmail defends us against more effectively than Protonmail?
On the other hand, they have more resources than anyone else to protect against things like DDOS, nation-state hacking/phishing, and physical disasters. They also have a legion of lawyers to protect against improper legal requests, however they will roll right over for a government if it's legal.
Protonmail is on point with the privacy, but their security engineering team is probably less than 1/10th that of Google's.
https://blog.google/products/gmail/g-suite-gains-traction-in...
They are definitely still reading your gmail. How else would the spam and other filters work? They can also use it under this policy for anything but "ads personalization". Machine learning, Google product integration, other recommendation not deemed to be ads, refining your google profile, etc. A very narrow scope of exclusion.
Spearphishing is a huge source of compromise at the moment; antiphishing filters might, in that view, be considered a security feature rather than a security fault.
On the other hand, I have, frankly, never understood these privacy arguments. Is it a privacy violation if someone checks a checksum of my incoming mail against a blacklist? What if they compute a hash of my mail to check the DKIM signature? And if those are OK, why is an ML model more of a problem?
No, the privacy concern is about the potential for abuse if there is a known place in the world where a very detailed account of all my online (and some offline) activities, contacts, communication and personal interests is stored.
It obviously arouses the interest and desires of criminals, governments, employers, politicians, landlords, insurance companies, creditors, marketers, ex partners, extremists and bigots, journalists, potential mates, researchers, etc.
Some of these groups are overlapping and some of the concerns might only arise in the distant future. Governments change. Ownership of companies changes. Personal circumstances and opinions change, but you can't take anything back once its out.
So even if someone hacks into the spam filters, they won't be able to reconstruct content from your account. They _might_ have insight into things like word usage, but they won't have reconstructed sentences or the like, a priori.
You can build systems like this without having long-term storage of the content, which protects against data issues for one-time leaks.
An email with the order went to my @gmail. The next day in my Youtube videos the ads where for stop smoking patches.
They read email.
Sometimes hilariously. A few years ago a friend emailed me a barbershop quartet video, and I spent the next couple of weeks getting hairdresser ads ...
And, if it's being interpreted by a machine, does that really count as reading?
One could argue that Google's filtering of spam and potentially harmful mail is a point in favor of Google.
Email headers contain a lot of information. It has the various email addresses, servers involved, ip addresses, time stamps, subject, priority, and things like that.
The body of the email is the only part that gets encrypted when encryption is in use.
My entire family uses Gmail (our domain is hosted via Google Apps) and as far as I know, it's been very close to perfect for all of us.
So team size alone doesn't convince me one way or the other, even if I were to completely disregard all privacy issues.
I am a Gmail user as is my company, so I do trust them quite a bit. But I feel that Google has a much bigger problem on its hands than Protonmail. Both because of its business model and because Gmail does things like search and spam filtering, which Protonmail cannot do.
[Edit] Protonmail actually does spam filtering.
Gmail + 0$ per month = zero privacy for you and anyone who emails you, plus Uncle Sam has full access to your life.
Protonmail + 4$ per month = you will never see ads for a <insert_item_name> like the one you just bought, plus you will be driving Uncle Sam crazy!
Encrypted messaging apps and services like ProtonMail have never been primarily to help people with Snowden's threat model. They're for people like you and me to reclaim a semblance of privacy, and they work even with "Uncle Sam" as the threat model in a limited, dragnet surveillance sense.
They don't work, because the US government's modus operandi is compromising machines or forcing users to provide access to their encrypted data. It's unclear to me why, if you take as premise a government capable of forcing one of the most valuable organizations in the world to hand over its data, you believe a company several orders of magnitude smaller is safe because it's "end to end encrypted" and has servers in Switzerland.
Put another way, I find the concept of a government willing to force Google to give up data but unwilling to use operational vulnerabilities to achieve the same thing to be contrived - how is this not just an arbitrary line in the sand?
Furthermore, the heuristic itself is a red herring, in my opinion. It is far more likely that Protonmail has a critical security vulnerability inherent to its software than Gmail does. And even if we assume that the government doesn't want to spend economic resources on actively compromising you as an individual, why would the government not spend resources on a system to compromise you passively as part of an en masse campaign? In other words, are you using a custom built computer with parts designed by a boutique firm from another country immune to the wiles of government backdoors?
How do you decide where you want to stop down the rabbit hole, and are you really doing so empirically?
In the US we have a constitution the prohibits searches of our papers without a warrant signed by a judge. It might be out of fashion is some circles, but the rule of law and not just rule of power is quite popular and I would say a superior system of governance. Many Chinese who are acquiring assets outside of China feel the same way.
"use operational vulnerabilities" am I the only one who strongly believes that Micro$oft is in bed with every 3-leter-agency in haning out backdoors/vulns for the last 20 years?
I'm not so sure - at least as recently as 2013, Lavabit showed that even top level US govt targets had some realistic reliance on properly encrypted 3rd party email providers...
The "dragnet" is the thing that's potentially useful - if it's difficult enough for them, they can't do warrantless "full take" surveillance - even for non US citizens, then choose to individually target you later based on a complete historical record being open to keyword/"selector" based searches.
(And for the appropriately paranoid - even Levison's comments back then suggested the thing he was prepared to fight and maybe go to jail for was handing over the SSL key that'd have exposes _all_ users. Reading it the right way suggests he may have sold Snowden out on his own - and I can't exactly say I wouldn't have done so myself in his position - but he was principled enough to not hand over the keys to the entire userbases's security. I sincerely hope _I_ never have the protection of privacy of a user like Snowden being my responsibility while the full pressure of the US government bears down on me. I strongly suspect my strongly-held personal principles would not stand up to that...)
ProtonMail doesn't meaningfully address the mass surveillance aspect, though. Most emails still hit its servers in plain-text form. Encrypting once it hits their server doesn't help the mass surveillance aspect, it only helps the targeted surveillance when a warrant comes in.
And if you're willing/able to get everyone that emails you to switch to PGP to get real end-to-end encryption then protonmail is worth even less, since none of their benefits matter anymore (google is obviously not able to decrypt your PGP emails, either).
For me - I think they're useful protecting against dragnet "full take" surveillance (especially since I'm a non-US citizen, so am considered "fair game" for warrantless surveillance), but I don't for a moment think they'll protect me from any sort of state actor level interest targeting me specifically (I'm still gonna get Mossad`ed upon...)
(In more paranoid moments, I suspect that the first "dragnet" protection quite probably makes the second "targeted interest in _me_" more likely...)
There’s a reason all my systems use other cards, and are behind a hardware firewall specifically configured for my use cases.
Let's assume a system where every piece of hardware has a closed device driver, or part of it, CPU included. We're there, or very close. It's not that hard to imagine a system within the system that can access data (hard drives have closed blobs), read passwords before they are encrypted through keylogging (USB sniffing), make screenshots of the desktop (video card closed blobs) and send them wherever they're instructed to (network card blobs), not to mention downloading and executing arbitrary code.
Now one could object that the traffic could be easily intercepted, but what if all network chipsets of all vendors, including those inside routers, had a small set of instructions to intercept any magic packet satisfying some rules and treat it differently. Let's say send it to some hardcoded addresses without counting them or reporting them to user applications; even leds on front panels would not report those packets passing through. The only way to realize something fishy is going on would be by tapping physically into the network cable using non-network dedicated chipsets, say very fast digital analyzers, decode all traffic and match it with what a normal sniffer would report.
I admit this is a crazy scenario, but if an entity with nearly infinite resources had the power to force any hardware vendor to put spying hardware/firmware into every machine, wouldn't it attempt to do something like that?
Putting my tinfoil hat on, I'd say that they got a new-improved way to get in, and they patch the old one because now the 'others' got whiff of this (e.g. shadowbrokers) and are about to start abusing it themselves :)
In other words, your comment is pointless.
https://www.usenix.org/system/files/1401_08-12_mickens.pdf
Threat: The Mossad doing Mossad things with your email account
Solution:
* Magical amulets?
* Fake your own death, move into a submarine?
* YOU’RE STILL GONNA BE MOSSAD’ED UPON
“The government can hack anyone, just give up” is a dumb objection if you view security from an economic perspective. Defenders have a huge advantage over attackers that we aren’t sufficiently taking advantage of yet.
I am not sure I understand the “fear” of the US government. Do we have cases of “normal” people being harmed from NSA type activities? We’ve had a ton of cases of normal people being harmed from non-government “hackers,” so, from a risk management perspective it seems silly to prioritize surveillance avoidance over garden variety thieves. To think Proton has the same level of experience and technology that Google has is a bit naïve.
And server location doesn’t particularly mean much. Plenty of Swiss banks have been compelled to turn over US citizen information due to FATCA — it’s not a stretch that a legitimate request for information by the US government would be honored by the Swiss if it pertains to a US citizen. For non-US citizens, there might be some benefit to an offshore server, however email is generally not the weakest link in surveillance. Also, you’d need to ensure that all your recipients are also using non-US as well as non-British, or non-French systems as well.
US surveillance is in the spotlight, but France and the UK are equally aggressive, if not more so since the actual laws in the UK and France are much more liberal in terms of allowing government to intercept communications. The French law doesn’t even require a judge (secret or not.)
https://www.recode.net/2015/11/14/11620670/france-has-a-powe...
And then there is this law In Switzerland— backed by almost 70% of voters:
http://www.bbc.com/news/world-europe-37465853
I think Proton is a nice alternative, but other than effective marketing, there isn’t much differentiation from paid Google Apps/Gmail plans. Comparing “free” gmail with Proton isn’t exactly honest, comparing paid gmail to Proton is probably a more valuable comparison.
With Gmail, for example, it’s possible to get a Business’s Associates Agreement for HIPAA compliance.. which means that it’s possible to have email that’s more secure than “normal” free Gmail. Of course HIPAA isn’t relevant to government surveillance, but really, how many people are actually at risk from the NSA? If that concern is part of your risk profile, then perhaps you ought to be living in a Tora Bora cave with messages being delivered encoded with a one time pad. If you are worried about your Antifa or KKK meeting minutes being intercepted, it’s likely Proton isn’t going to be much help.
First and foremost, despite widespread fetishization of things like end-to-end encryption, real world software rarely differentiates its security based on superior cryptography. Most security vulnerabilities occur at the endpoint level and are not even technical vulnerabilities, they're just successful, targeted phishing campaigns. The next most common set of vulnerabilities are in mundane software, typically in the infrastructure and peripheral logic. These are due to developer ignorance, misconfiguration or sometimes both. From there we have crypto implementation flaws, which are typically due to software engineers' fascination with implementing their own cryptography libraries or taking strange liberties with existing libraries that deviate from the explicit or implicit intentions of the original author. Finally, on a peak so remote and small it's scarcely visible from the ground, we have actual cryptography design flaws, where someone literally rolls their own crypto at the conceptual level and deploys it.
Second, email is a fundamentally antagonistic medium when it comes to end-to-end encryption. If you are actually concerned about your privacy and you're discussing something that warrants extreme care, email is the least user-friendly and most error-prone method of going about it. You should consider a synchronous medium with forward security if possible, probably based on a well-known and well-audited messaging protocol (to avoid a flame war I'm not going to suggest any particular one - do your research).
Third, if your threat model is honestly the US government, you need to significantly revise your opsec entirely if you're realistically considering Protonmail. Regardless of its actual security, you're (implicitly) saying that you trust Protonmail to be capable of withstanding the resources of a motivated three letter agency with an armada of security compromising tools at its disposal. Why trust a third party at all then? What makes you think the servers being in Switzerland is going to help you if you distrust the government this much? By all means, don't use Gmail either, but then Protonmail isn't really a coherent security measure either - again, use a synchronous messaging platform, or develop the opsec needed to consistently use PGP correctly on your own.
Given the foregoing, if your adversary is actually the US government, neither Gmail or Protonmail are effective strategies, and if your adversary is not the US government, Google's security team is vastly more qualified and has overwhelmingly more resources at its disposal to secure its email infrastructure.
Plus, you could say they are describing a threat model. If ProtonMail were compromised in this one particular way the confidentiality of your mail would be 'stronger' or 'improved'. This should be as reassuring as 'Switzerland', which is, of course, also trotted out.
Without a threat model (that is, the set of threats that one is trying to secure a system against), you have no idea what someone means by "secure". It could mean unpickable doorlocks, it could mean unbreakable windows, it could mean angry-Hippopotamus-proofing. It could mean that you smelly farts can't escape your pants.
Any claim of security without a threat-model is in the most literal sense meaningless. And don't get me started on "Military Grade Encryption", which is a term that at this point should give you a sense of concern, rather than safety.
I'm not sure I understand what your counterpoint here is since we seem to be saying the same thing but I had to go back and fix 'thread model' twice myself. An underestimated threat model to commenting about threat models!
Calling this fundamental difference in approach "more secure" manipulates the less-informed instead of educating and almost eliminates the chance of a worthwhile conversation about tradeoffs and values that could be very flattering to ProtonMail.