HN Explorer

Gmail marked my email to my wife as suspicious?

9 points by selfhoster ↗ HN
I forwarded an email from a highly reputable organization, from my email account on a server I've run for 30 years successfully with high reputation, to my wife who has no issues with her Gmail account and GMail in their infinite either idiocy or insolence marked it as suspicious, scaring her.

Meanwhile, spammers set up temporary GMail accounts and blast spam into my inbox with fake McAfee or other fake subscriptions several times per week. The same patterns in the emails going on for months now, maybe over a year, regardless how many times I take the time to go to their site and fill out their form since they ignore their listed abuse address; their AI apparently isn't smart enough to spot these spammers where even a 6 year old would be able to spot them. I keep GMail whitelisted but wondering if I should remove them from my whitelist and start reading up on degoogling.

And they are a multi-trillion dollar organization? It doesn't add up.

14 comments

[ 3.0 ms ] story [ 46.0 ms ] thread
JojoFatsani
Bro don’t take it personally
CamperBob2
The idea that a spam filter doesn't immediately check to see if the sender is in the recipient's address book when judging ham versus spam has always seemed just nuts to me. Hello? If I've emailed someone before, don't ever mark email from them as spam!

Even the SpamBayes plugin for Outlook wasn't smart enough to do that, IIRC. And it definitely had access to the address book.

jeffbee
Sender-in-addressbook isn't a saving throw for malware.
CamperBob2
So? It is for false spam positives, which is what I was referring to. GMail cheerfully marks some of my email to my GF as spam, after 10+ years. Utterly inexplicable.

(Edit: not the OP, just sympathetic to his plight)

dghlsakjg
Did they mark it as spam or as suspicious?

OP said it was marked as suspicious. Spam and suspicious are not the same thing.

It is entirely understandable that they might have marked something as suspicious. Compromised email accounts from known senders are a super popular way to scam people.

tetromino_
Counterexample: I used to have a family friend who liked to send out fringe/conspiracy/general insanity political emails. Occasionally multiple ones per day.

Sure they were in my address book, but 75% of what they sent me needed to go straight to the spam folder.

CamperBob2
Eh, that's the edgiest of edge cases, not something that should drive default functionality.

Also easily addressed by manually marking the email as spam. The filter could treat that as a cue that the sender should no longer benefit from the address-book check. Better yet, store the flag as a visible field in the address book, so that you can turn it on and off yourself when you edit their contact information.

ratg13
“Server with high reputation” doesn’t mean anything.

DKIM, SPF, DNSSEC are things that matter

You could be doing things poorly for 30 years for all we know.

selfhoster
For the record, yes I have all of those and DMARC in place. My domain was literally registered with Google Domains so they had full access to all traffic through them before they made the unusual decision to sell Google Domains to Squarespace.
igor47
Yeah it's pretty frustrating, especially since they don't tell you. Gmail effectively is email, and I'm constantly paranoid my email is not being delivered. I too have been running my server for decades, and I do have spf/dkim.
vednig
Some corps are intelligent enough in utilizing their AI where they want and ensuring that problems like this happen, to indirectly maintain confidence in people that their data is not used for AI training and some other things.

They can do it better but choose not to because:

1. It's not their direct form of revenue (usually it's ads, so when it comes to recommending ads they could do a far better job at it than others)

2. They're not completely consumer focussed( at the scale they're playing at, the consumers don't matter, only features on papers do, that said, they still apply significant effort to maintain better customer experience, ps: just don't contact their support)

3.They know spammers how spams get by(because spammers evolve and so does their algo to detect spammers, and so on it becomes increasingly complex to identify spammers or bots, from real people, it's a recursion issue just like the concepts of virus and anti-virus, if there is virus,it will be blocked by AV but then there will be new virus and so on, loop never closes)

metalman
"scaring her" , meaning they have her trust. And there in lies the success that they are weaponising for further market share. The well established trend for each organisation to label ALL external comunications as untrustworthy, with long attachments detailing the precautionary measures and protocall before opening an outside message is only getting worse. The incentive to create "spam" as a false flag operation, is far to high, to dismiss out of hand. The end goal is of course to privatise all coumication mediums, to the point that fonts are private property, and for the illiterate there are "emojies" another privatly owned and censored medium, which of course are now convieniently availble for inclusionin little selfie vids, along with any number of other, non verbal, non literate,embelishments. So scare them a bit, and then provide nice safe sparkles. Personaly, I dont believe that any of these shenanigans cant reach there hoped for end states, as basic societal needs are dependent on an indipendent and impartial, mails and comunication system. It's still very irksome haveing to watch the mindless grasping for total power and control.
theanonymousone
One Gmail flagged an Email from Google as suspicious for me. Beat that :)