[HN Gopher] Gmail E2E is as terrible as expected
       ___________________________________________________________________
        
       Gmail E2E is as terrible as expected
        
       Author : HypnoticOcelot
       Score  : 149 points
       Date   : 2025-04-06 19:42 UTC (3 hours ago)
        
 (HTM) web link (michal.sapka.pl)
 (TXT) w3m dump (michal.sapka.pl)
        
       | ranger_danger wrote:
       | I think one of the growing threats lately in the community has
       | been over malicious client-side javascript, especially when the
       | client handles end-to-end encrypted content (used on sites like
       | Proton, MEGA etc.), so requiring users to trust Google with the
       | contents of these client pages, and by extension the emails
       | themselves, seems (in my opinion) to defeat the entire point of
       | this feature.
       | 
       | Some work in this area has been done in the form of browser
       | extensions that are used to verify signed assets delivered to the
       | client:
       | 
       | https://github.com/freedomofpress/webcat
       | 
       | https://github.com/tasn/webext-signed-pages
       | 
       | https://github.com/jahed/webverify
       | 
       | https://github.com/facebookincubator/meta-code-verify
       | 
       | But unfortunately for now, none of these are seeing wide adoption
       | and this remains an unsolved issue. It also does not require
       | anyone to use known-good, audited and verified open-source
       | components, meaning even if the client code is signed, it can
       | still be malicious... there must be a greater reason to trust the
       | code than just "trust me bro".
        
       | kardianos wrote:
       | This is how all HIPAA "secure email" works. Outlook, Zoho, clinic
       | comms, BECAUSE it lets you revoke email access.
       | 
       | If you want an opportunity in this space, it isn't actually
       | encrypted emails, but possibly standardizing and streamlining
       | such "message pointers" and address endpoint verification.
        
         | perching_aix wrote:
         | > lets you revoke (...) access
         | 
         | That's pretty funny.
        
       | zer0zzz wrote:
       | Doesn't proton mail also do this for sending encrypted mail to
       | folks with no encryption?
       | 
       | They ought to do pgp though
        
         | jitl wrote:
         | Can't expect a civilian to manage pgp keys or go to key signing
         | parties
        
           | XorNot wrote:
           | The anti-establishment fervor of open source crypto
           | developers is the reason this is a problem though.
           | 
           | Most people, for most things, don't need to verify trust
           | outside of normal government channels.
           | 
           | i.e. any business I correspond with, trust is via the
           | government that they are a business bound by the relevant
           | legal system I live in.
           | 
           | Same story with communicating with basically anyone: if their
           | GPG key was signed by the common government key, then hey,
           | good enough for anyone.
           | 
           | The problem is...we don't have the infrastructure for any of
           | this. And GPG key servers are inadequate for maintaining
           | suitable privacy for people if they were used at this scale.
           | 
           | But we certainly could provide the means by existing
           | technologies: e.g. nothing stops us making drivers licenses
           | and other forms of ID smart cards.
        
             | solardev wrote:
             | It depends a lot on the government in question too. In the
             | US very few people would trust the government to handle
             | something like this. The certificate authority system is
             | run by big tech companies and nonprofits, for example, not
             | the government.
             | 
             | Trusting the government as a peer makes sense for
             | government sites, but for anything else, it just makes
             | censorship way too easy.
             | 
             | Even among businesses, we normally trust the middleman (the
             | credit card issuer, and their protections and chargebacks)
             | over the government. If a business screws over a regular
             | consumer, the government isn't really going to do anything.
             | 
             | Maybe you have a more civilized society and functional
             | government where you live. We don't.
        
               | XorNot wrote:
               | You've just listed out a bunch of alternate trust roots
               | which, with appropriate infrastructure, could also easily
               | provide practically secured E2E communications which
               | would be adequate for real world use.
               | 
               | The point isn't "trust the government" the point is trust
               | is contextual and the notion of "key signing parties"
               | always missed it (and if you actually go and read up on
               | the concept, government ID documents were considered to
               | be something to ground whether a signature should be
               | issued by someone so it was already baked into the system
               | anyway).
        
               | solardev wrote:
               | Well, I think that's actually the point, no? This was
               | never a technical challenge (not for a few decades,
               | anyway), but a question of which authority to trust.
               | 
               | Companies rolled their own out of a commercial need. The
               | FOSS community didn't trust the government or big
               | companies so never fully solved the problem. Users just
               | don't care.
        
       | Freak_NL wrote:
       | This was already happening, unfortunately. The user's mail agent
       | is deemed untrustworthy (and so is the user), so every service
       | which needs to send confidential data just turns your email into
       | a notification with a link. There are so many of these, but often
       | they are limited in scope. For sectors like healthcare you have
       | companies offering this type of service to companies which need
       | to adhere to security theatre standards such as ISO 27001, and
       | because nations often have their own added requirements for
       | specific sectors (think HIPAA in the US or NEN 7510 in the
       | Netherlands) these services tend to remain focussed on single
       | countries.
       | 
       | Then there are the national governments and things like insurance
       | companies. All happily sending message notifications where you
       | need to sign in to their own portals.
       | 
       | Securing email is too complex, so everyone builds their own
       | secured portal thingy, and your mailbox has become a receptacle
       | for notifications. Figuring out a solution would require
       | cooperation, pragmatic lawmaking, and giving up those nice
       | cashcows of moated portals, so it won't happen.
        
         | modeless wrote:
         | I hate this so much. I will pay for a service that takes all
         | these content-free messages and goes to the website and logs in
         | and extracts the actual message content and puts it in my
         | inbox. Anyone want to make that?
         | 
         | I actually think there is a more general opportunity here with
         | AI. Every app and website and UI I use is optimized by a gaggle
         | of PMs to achieve business objectives that don't necessarily
         | benefit me. AI is getting to the point where soon it will be
         | able to use these non-aligned UIs for me, and present to me a
         | much simpler UI customized just for me, that does what I
         | actually want and no more.
        
           | jack0813 wrote:
           | You can probably do this fairly easily already, like this:
           | 
           | - Fetch your emails using any of the common local mail sync
           | tools
           | 
           | - Some processing to clean up the plaintext version, may not
           | be necessary even
           | 
           | - Send it to an LLM to extract a link
           | 
           | - Open up a headless browser, trigger something like
           | SingleFile to extract its content.
           | 
           | Though you'll have to keep the cookie refreshed, but if it is
           | initially logged in, this should be fine since you can also
           | program something to keep refreshing every once in a while.
        
         | smileybarry wrote:
         | Yep, while I like moving all snail mail to email, I hate that
         | (almost) every single service now just sends a
         | "monthly/quarterly report now available in your account area!"
         | email. A rare few of them at least offer the option of just
         | sending it attached (with the default being a useless reminder
         | email), but most are essentially a chore, because nearly
         | everything here uses phone+2FA as login rather than a password
         | or passkey.
        
           | LexGray wrote:
           | Passkey does seem like it should include functionality to
           | encrypt and sign all email and SMS automatically. A missed
           | opportunity.
        
         | bob1029 wrote:
         | I struggle with how the secure email solutions are inherently
         | more secure than just dumping the pdf or ticket details in the
         | email body.
         | 
         | Every vendor's secure email portal I have ever used was
         | ultimately authenticated using my email account. Any one-time
         | passcodes are sent to the same email. Password recovery? Email.
         | If a malicious user is on my PC or otherwise intercepting my
         | mail, they could access 100% of the solutions I've got access
         | to right now.
        
           | dvdkon wrote:
           | I don't think that's true for my bank or Czech government
           | services. Plenty others do practice "security by email",
           | though.
        
           | bigthymer wrote:
           | I always understood it as, the email with link notification
           | thing started as soon as email providers began regularly
           | scanning users' emails. Before then, an email included all
           | the information you needed without having to login to another
           | site.
        
         | irq-1 wrote:
         | > The user's mail agent is deemed untrustworthy (and so is the
         | user)
         | 
         | Bluesky follows this pattern for the benefit of the user. The
         | internet tradition has always been: if you want to control it,
         | you have to host it.
        
           | Dylan16807 wrote:
           | > if you want to control it
           | 
           | In the context of sending a secure message, the sender
           | maintaining control goes in the _negatives_ column. At best
           | it 's a compromise in exchange for specific security
           | features.
        
         | tshaddox wrote:
         | Not that I love any of those particular secure document web
         | services, but I do vastly prefer web pages and URLs for
         | retrieving documents and would happily live in a world where
         | email was a receptacle for notifications.
        
       | solardev wrote:
       | Isn't this how banks send secure messages too? You can't send
       | secure emails to arbitrary clients otherwise. PGP stands no
       | chance of being adopted by regular users.
        
       | stwrzn wrote:
       | What happens if the sender's Google account ceases to exist for
       | whatever reason? What if Google ceases to exist?
       | 
       | I know that there are a lot of HIPAA "secure email" solutions
       | that also do this, but I don't want this to become more common
       | practice then it already is...
        
         | toomuchtodo wrote:
         | Keep HARs from the browser if you need your own record of the
         | messages. Next step would be to determine how to extract the
         | messages from a history of HARs and inject into your own
         | mailbox or other storage system for archiving and search.
         | Perhaps a browser extension to automate this automated logging
         | of message retrieval.
        
         | solardev wrote:
         | IMO those are different use cases. If that sender or Google
         | itself were to disappear, hopefully the messages would just
         | disappear too. It's better that they become inaccessible rather
         | than public.
         | 
         | Long term archival is a different use case altogether,
         | especially of encrypted materials. It's questionable whether
         | any provider or medium can survive over the long term, so it's
         | better to use an encryption system where you hold the keys and
         | the encrypted data can be migrated to any sort of storage or
         | provider over the years.
        
         | TJSomething wrote:
         | I just keep all my Gmail synced into local Thunderbird via
         | IMAP.
        
       | n3storm wrote:
       | I guess minimal gmail for e2e is the tariff we have to pay for
       | privacy and freedom. wink.
        
       | otterley wrote:
       | > I'm not going into how much this is not E2E, as this has
       | already been proven
       | 
       | By whom? It looks like E2E to me. It's just that both ends are
       | controlled by the same entity.
        
       | p2detar wrote:
       | The E2E problem has already been solved long time ago. We used to
       | have Thunderbird with an OpenPGP extension and GPG keys.
       | 
       | Then there were a whole plethora of products were build around
       | Lotus Notes Domino that provided a central place for securing
       | outgoing E-mail using either S/MIME or GPG keys. All of this on
       | premises. Then came the Cloud and obliterated these products. And
       | for what?
       | 
       | edit: typos
        
         | bigfatkitten wrote:
         | > The E2E problem has already been solved long time ago. We
         | used to have Thunderbird with an OpenPGP extension and GPG
         | keys.
         | 
         | It was never really solved, PGP email is a usability disaster.
         | 
         | Client support (especially on mobile) is limited. Headers,
         | including the subject line remain in cleartext. Users forget to
         | click the "encrypt email" button and so messages go out in the
         | clear; sometimes in reply, and so the entire conversation is
         | exposed. Key exchange with new recipients is tedious to do
         | securely.
         | 
         | Not to mention the extra issues caused by HTML in message
         | bodies.
         | 
         | https://www.eff.org/deeplinks/2018/05/not-so-pretty-what-you...
        
           | AshamedCaptain wrote:
           | Ah, the usual "A is an unpopular standard, so we decided to
           | do our own proprietary thing". Guess what: for it become
           | popular, it actually has to be implemented. If Google
           | implemented it in Gmail, many of these purported issues would
           | go away . It would still not truly be E2EE mail, but then
           | again neither is this monstrosity.
        
           | aborsy wrote:
           | It's a disaster because email providers don't want to offer
           | E2EE or make it easy.
           | 
           | Is it that hard to generate a certificate for each email
           | address client side and store that, and the private key
           | encrypted with the user's password, on the provider's server?
           | 
           | The majority of email is gmail and Google could make that
           | E2EE by default.
           | 
           | Countless products that have successfully implemented public
           | key distribution (proton mail, instant messaging, ...).
        
             | pas wrote:
             | that's not the hard part. it's the out-of-band key
             | exchange. (or key discovery/verification. so basically how
             | to avoid the trust on first time use problem.)
        
           | Telemakhos wrote:
           | While I'm sure some of these flaws also apply to S/MIME, I
           | feel like its client support (even in Apple iPhone native
           | mail app) is far superior to PGP. Apple made S/MIME
           | installation and use across its ecosystem, and I remember it
           | being easy in kMail once upon a time when I used KDE; why
           | didn't S/MIME ever catch on?
        
         | josephg wrote:
         | For PGP keys to work, users also need to publicly broadcast who
         | they know and personally trust. It's a privacy disaster.
        
       | tptacek wrote:
       | This is how most commercial secure email products work.
        
       | fnord77 wrote:
       | how else are they going to push ads?
        
       | jonathantf2 wrote:
       | This is exactly how M365's solution works. Decrypts if the
       | recipient is using MSN or EXO, otherwise punt them to a webpage.
        
       | jurschreuder wrote:
       | People in China could not open a url sent in Gmail. I happened to
       | be in China, I tried to open the webpage and it worked, no
       | firewall.
       | 
       | I hovered on the link in Gmail and Chrome told me left bottom it
       | was just that exact url.
       | 
       | But when I opened the url it got blocked by the great firewall.
       | 
       | Why? Any link in Gmail secretly gets replaced by a link to Google
       | that tracks you and then redirects you to the original link.
       | 
       | The most obnoxious thing about this I think is that Chrome shows
       | the original link.
        
         | josephg wrote:
         | That's a classic technique to track click through rates.
         | Google.com has done this forever. The technique is to make an
         | actual link in HTML (<a href=...>) then add an event handler
         | which cancels the link's default behaviour when you click it -
         | and replaces it with javascript, or a tracking link.
         | 
         | I understand why Google.com wants that data. But in an email
         | client it's extremely obnoxious.
        
           | ChadNauseam wrote:
           | If it's replaced with a tracking link, I think it might be
           | just as effective to use the `ping` property on browsers that
           | support it
        
           | SparkyMcUnicorn wrote:
           | I thought Gmail didn't support js execution. Did Google make
           | an exception for themselves?
        
         | mzajc wrote:
         | > The most obnoxious thing about this I think is that Chrome
         | shows the original link.
         | 
         | Not defending Google here, but could it be that the tracking
         | page is opened by onclick() while the browser shows the
         | original href attribute?
         | 
         | I've seen some websites do that for various purposes and
         | opening the link in a new tab (middle mouse click) usually
         | bypasses it.
        
       | tky wrote:
       | All the takes on this release miss one crucial point: if you want
       | people to adopt E2E encryption, you must reduce friction. For
       | users of Gmail, that means familiar elements and flow to their
       | usual use of Gmail. If this lets even a handful of people use
       | more secure messaging, it's a win. For Google workspace-centric
       | orgs it's a good step in the right direction.
       | 
       | If you disagree, go set up GPG on a non-tech's computer, tell
       | them they need to use Thunderbird or some other helper app(s),
       | and see if you can even go home before being asked to remove it
       | all.
        
       ___________________________________________________________________
       (page generated 2025-04-06 23:01 UTC)