How does blocking pictures in an email protect my privacy?

Email programs commonly block remote images. I'll look at why, and how spammers and others use remote images for good and evil.

//

In Windows Mail, I received an email from a known vendor (not spam) with all the pictures withheld. At the top (below the header) there was a message which read:

“Some pictures have been blocked to help prevent the sender from identifying your computer. Click here to download pictures.”

My question is: How can a sender identify my computer by me receiving pictures? And of course, how great is the risk?

“Identifying your computer” in that informational message is somewhat vague, as it’s not exactly what can happen. But the concept is still important.

And in fact, if you’ve ever seen ads or services that claim “we can tell you if your email has been read” – images are how they do it, and it’s also why they can never be 100% reliable.

Email image types

There are two types of images in email: attached and remote.

Attached images are exactly that: they’re attached to and sent with the email message itself. When the message is displayed, your email program need look no further than your own computer for the images, and typically will display them right away. Attached images can be displayed in the body of your email, or they can appear as actual attachments. In either case, however, they were sent with the email itself.

Embedded/Attached Image

Email with an embedded or attached image. Image is included with the message.

Remote images, not surprisingly, are also exactly that: remote. The email message actually includes a link (or reference) to the image, rather than the image itself. The intent is that when the email is displayed, the email program will use the link to fetch the image from wherever it resides on the internet, and display it.

Email With Remote Image

Message with a remote image: it is fetched from the email server only when the message is displayed.

In either case, the email can look exactly the same.

Each of the two techniques has pros and cons. Attached images make your email larger and slower to download and deliver, but probably faster to display. Remote images create smaller emails, but assume that the images can be fetched at the time the email is displayed, which is not always a valid assumption.

Most email programs today will block remote images by default, unless you indicate that a particular sender is “safe” by adding it to an address book or some kind of safe list. This is typically what’s happened when an email arrives and all you see are red dots or empty holes where the images should be.

Now that we know what remote images are, let’s look at how they can be used and abused.

How remote images work

Let’s say I send an email to a large number of people. In that email – formatted in HTML – I include a link to an image on my server, like this:

https://img.askleomedia.com/puppy200.jpg

In HTML, it would be encoded like this:

<img
  src="https://img.askleomedia.com/puppy200.jpg"
  alt="Corgi Puppy!" />

And if displayed, would look like this.
Corgi Puppy!

That image is hosted on my web server1, and because of that, each time the image is displayed by an email program, or by someone reading this article, the retrieval of that image is logged.

That’s how I can track how many people displayed the image. Note that this is not the same as the number of people that opened the email – since images are blocked by default, there’s no reliable way to do that.

So far, though, we haven’t identified anything about you specifically. That doesn’t take much of a leap, though.

Tracking you via remote images

Let’s assume once again that I’m sending a large mailing to a number of people. In that email, I include the same image, but this time I add a little something to the HTML:

<img 
  src="https://img.askleomedia.com/puppy200.jpg?email=you@example.com"
  alt="Corgi Puppy!" />

Here the mailing list software automatically includes the email address the message is being sent to as part of the link to the image.

There are many, many techniques for doing and hiding this, but I’ve chosen something easy and obvious for this example. Each email contains a link to the image, “personalized” with the email address of the recipient.

When the email program fetches the image, it asks for:

http://ask-leo.com/images/askleonew.png?email=you@example.com

Since this is an image, everything including and after the question mark is ignored by the web server, which just returns the image.

But the full reference, including the “email=” part, is logged in the web server’s access logs.

And now I can tell not only that the message was opened and the image displayed, but exactly who opened it.

Unless, of course, images are blocked in your email program.

Abusing remote images

So how can this information be used and abused?

Well, with spammers it’s easy: if they see that a particular email address has opened a message and displayed an image, the spammers now know that they have a valid email address that someone actually looks at. You can expect more spam.

With legitimate businesses it’s a little less clear, and in my opinion, a lot less concerning. They do track “open rates” to see how effective their messages are. They know that not everyone displays images, but they can observe trends in the portion that do.

Businesses can also track individual open rates if they want to, using the techniques above. Exactly how they might use this information varies, depending on the business. As one example, I use it to see how many people open The Ask Leo! Newsletter. Among other things, this helps me understand just how useful the newsletter is to my readers.

My configuration is simple: I leave images blocked by default, but add almost all the business senders to my whitelist, so I can see their messages in all their glory.

Play
Footnotes and references

1: Actually on a CDN – a Content Delivery Network – which hosts the image on their servers around the globe so that it’ll be closer to you, and therefore faster to load. But the concepts all still apply.

Comments

  1. Mark Brownlow

    Great explanation.

    In theory, legitimate business emailers should be using open rates to make the emails they send you more relevant. To give an example:

    Say you signed up for an email newsletter with information on cruises, because next year you want to go on one up the Amazon.

    The emails the travel company sends you are sometimes about European cruises and sometimes about South American cruises.

    If the company sees that you only “open” those emails about South American cruises, then they should change the focus of the newsletters you get. So there’s more on South America and less on Europe.

    In reality, though, most businesses don’t get that clever and simply look at how many people in total open their emails so they can get a general feel for want people like most.

  2. Greg Bulmash

    I’ve heard a number of terms for these tracking images. Some call them “web bugs,” but my favorite is “ratware”.

    I do agree with Mark that it would be nice if open rates were used to personalize and target e-mail ads better, but they’re just generally used to determine the success of a specific campaign, and at best they try to use the information to make their future subject lines more clickable to their broader mailing list.

  3. George Arauz

    It actually doesnt protect it… maybe prevents a trojan virus here and there… but my advice.. dont open any emails you dont recognize.

  4. susan shaw

    I recieve emails from stores to purchase items and you always need to click on them to see or by but my hotmail won’t allow me to open them. I have always opened them in the past but sense I changed messenger 3 days ago, it wont let me open anything.What can I do (in simple terms please)? Thank you susan

  5. Dan

    I’ve heard that images in emails can be used for even more nefarious purposes, like secretly downloading malware onto your computer when you open the email containing them. Can anyone comment on that?

    • This is not true in general. Once upon a time there were bugs (aka vulnerabilities) in various vendors image display code, but there’s no currently known vulnerability. The act of displaying an image is generally safe. It’s fetching the image from the remote server, as discussed above, that has privacy ramifications.

Leave a reply:

Before commenting please:

  • Read the article. Seriously. You'd be shocked at how many people make comments that prove they didn't.
  • Comment only on the article. If you have a new, unrelated question start with the search box at the top of the page.
  • Don't post personal information. Email addresses, phone numbers and such will be removed.

VERY IMPORTANT: because of a rise in comment spam that's making it through our filters any comments that do not add to the discussion - typically off topic or content-free comments - run a very high risk of being flagged as spam and removed.

If you have a new question unrelated to the article above, ask it on the Ask Leo! ask-a-question page.

Your email address will not be published. Required fields are marked *