Is there a way to send a response back to unwanted email that looks like it is from the server saying the email address doesn’t exist? I keep marking stuff as spam but of course they’ll keep sending it. I would imagine that even relentless spammers cull their lists for non-existent addresses.
A)Yes*
B)It’s not worth your time because, at best, it won’t stop and at worst it will confirm there someone checking the email. Spam servers don’t care, at all, if they’re sending spam to non-working addresses. The only way this would have any impact is if you’re not talking about automated spam but actual people sending you email you don’t want.
*Set up a filter that, upon getting junk mail (which you’d have to define either by address or keywords etc), it’ll reply with whatever you want.
ETA, when I get spam in cases where I know someone is reading the replies (like junk from car dealers), I’ll reply with “TAKE ME OFF YOU JUNK MAIL LIST” or something along those lines, but I’ll carbon copy their address a few dozen or a hundred times. If I can find email addresses for anyone else at the dealership, I’ll CC them a bunch of times as well.
I’ve since learned to use my double secret spam email account, the one I only even check once a week or so, for any interactions with dealerships.
Thanks–I could send autoreplies but I would want one that looks like it’s coming from the SMTP server, not me. But if they don’t care about non-working addresses then it’s a fool’s errand.
That’s not a particularly helpful response. Generally speaking, you can’t use a garden-variety e-mail client to convincingly craft a message that resembles a bona fide mail server bounce message.
No, it won’t stop all spam, but it will stop some. Buyers of spam mailing lists put a premium on ones with working e-mail addresses. So purveyors of such lists have an incentive to remove non-working addresses.
Only if you use your method. A bounce message from the mail server (not the client) is pretty definitive proof that the e-mail didn’t get read by a human.
If this were true, they would be generating e-mail addresses entirely randomly. In fact, there is a market for “verified” e-mail lists. You might know this if you read any of your spam, which is one of the main ways of advertising such lists. 
To the OP: What you are proposing is possible, and may have some small but measurable effect on the volume of the spam you receive, but to do it properly you will probably need enough technical know-how to run and properly configure your own mail server. Even if you have such knowledge, the effort is probably not worth the benefit.
A slightly less onerous strategy (but one with similarly low success rates) would be to set up your e-mail client to automatically forward spam to reporting services such as collaborative blacklists and SpamCop. Collaborative blacklists keep track of where spam comes from, and then distribute this information to users of the blacklists to use in their spam filters. (Your e-mail client might already make use of such a blacklist.) SpamCop parses the header and and links contained in your spam and sends reports to the people running the responsible mail servers or spamvertized websites (who are usually, but not always, ignorant of the spammers’ abuse of their systems) so that they can shut the spammers down. But most operators don’t take action, and even for those that do, the spammers will simply move on to other hosts. It’s a never-ending game of Whack-A-Mole.
Just copy one. Send an email to a fake address and copy/paste the message you get back. Granted, if a robot is reading/deciphering the reply, it may not work, but if a human sees it, they’re not likely to pick through the headers/body of it.
The absolutely do send emails to random and unverified emails. I know they pay more for good ones, but they have to find the good ones somehow. There’s plenty of stories out there from people that got spam before they even used their address. At my work I get tons of spam on email addresses that have never, ever been used in a form and don’t even have the ability to generate outgoing mail. Also, there’s the time that I got something like thirty thousand spam emails. All of them were bounced emails sent to (presumably) bad addresses and, for whatever reason, the spammer used my email address as the return path. I ended up having to call my email provider and have them turn off my account for a few hours and delete the messages from their because Thunderbird/Outlook couldn’t handle it and kept crashing my computer.
I’m not saying a fake bounced reply is a perfect (or even good) way to do this, it’s just A way to do it. I’ll still say that the best option is to simply delete (even if a filter is used) all those emails. If there was some magic bullet to stop spam, such as sending a bounce reply, everyone would be doing it. In fact, it many email providers/ISPs would likely automate it.
I did say “convincingly”. Copying and pasting an existing bounce message is not very convincing, neither to a robot nor to a human who knows how to read e-mail message headers.
Again, you’re ignoring a key operative word here, which is “entirely”. No spammer generates e-mail addresses entirely at random. At most they will randomly generate the usernames for a set of known domains.
And this is exactly my point: if n% of spammers make a point of using verified lists, and you can fool these spammers into thinking that your e-mail address is non-working, then you will reduce your spam by n%.
ok
It’s a Catch-22 basically. If a bulk emailer has their act together enough to accept and process bounces, they are probably honorable enough to actually Unsubscribe you when you click the link.
I’m not so sure. Even if sending bulk emails is relatively cheap, it’s not free. If they know that a percentage of email addresses are no good they can make more efficient use of their resources if they get rid of them. However, an Unsubscribe click says, “Hey, this is a real email address that a human is reading!”
There is an email client called IncrediMail that has a fake-bounce feature. However, I have no personal experience with this product, so I can’t tell you how well it works.
I suspect it wouldn’t faze spammers at all; it’s probably more useful for situations like making your ex think that your email address doesn’t work anymore. 
A lot of spam is sent by distributed networks of malware-infected machines; there’s not even a meaningful return path for a bounce message
My god after all these years of fighting spammers, why can’t we just have ISPs charge 1c per sent email. I’d gladly pay and that would put a crimp in the spammers who send hundreds of thousands of these annoyances.
I run a list server for a hobby (the subject is a particular hobby, and running it is my hobby) that handles ~5 messages/day to ~200 people. Entirely opt-in through hobby sites, carefully managed, unsubscribes are respected, etc. It’s been running for nearly 20 years. I couldn’t spend $10/day to keep it going (or spend even more in my time to collect fees from the userbase).
In any case, that would only work if every ISP started charging for emails. Even if most of the ISPs in the world got together and agreed to do this, there would still some ISP in India or Russia or whatever which would still offer free email. And they make a pretty penny since all the spammers would start using them.
No one involved in the spam business has humans reading individual responses and acting on them.
Professional spammers know a lot more about the technical details of how email works than probably anyone in this thread. There is zero chance that this will work.
That is the reason that sending fake bounces is generally considered a bad thing. They are almost certainly going to an innocent third party, if they go anywhere at all.
It is possible to setup procmail so that messages are bounced as if the address doesn’t exist (or any other failure mode, such as out of space). The main problems with that are: you probably aren’t running procmail; and as above, by the time procmail runs the message has already been received so the bounce is likely to go to the wrong place.
The way I handle it is by banning incoming addresses at the SMTP level. Then the email is rejected as it is being delivered, so whatever server (a bot; a legit, but annoying, server; etc.) is attempting the delivery will have to handle the failure. A bot is likely to do nothing and move on. The annoying car dealership will get an “error: nouser” and likely stop bugging me.
If spam is still a major problem for you, then you need to think about mitigation and filtering at an early place than your inbox. That might mean using a different email provider, if you’re using one that just can’t get their act together. Spam is still generally a huge problem, but at the mail provider level, for most people they are likely to see very few messages landing in their inbox.
That is for “real spam,” as the term has come to describe both unsolicited, probably fraudulent, advertisement, and legitimate email lists that people no longer want, or never actually signed up for, but were added to when doing business with somebody.