Our Dangerous Reliance On Email Attachments And What To Do About It

Updated: Mar 3

Continuing to send email attachments is costly, dangerous, and unnecessary, given the readily available alternatives. It is time to tackle the root causes of our cybersecurity challenges rather than the symptoms.

original post on Medium

Seemingly innocuous, email attachments are not. A design that dates back 50 years, inefficient from inception, has now become a destructive force diverting our precious resources as we continually tackle their collateral effects in collective resignation of a status quo that needn’t be.

Around 306 billion emails are sent and received every day. In the corporate environment, about 25% of messages carry file attachments, or 76 billion messages [1]. These are big numbers. Among the biggest, when it comes to any communication today. It is past time to question this most common file sharing method, especially when considering the billions of dollars spent trying to mitigate problems that find their roots in this outdated technology. An astounding amount of resources (time, money, effort, and lost opportunity) are directed in an attempt to secure, govern, and manage the detritus of this archaic model of file exchange [2]. It is time to question, in particular, because for years, secure file links have stood as a readily available alternative.

The challenge of email attachments stems largely from two aspects of its architecture, namely, duplication (data sprawl) of content and the lack of security.

Email is data sprawl on an epic scale

Data Sprawl

The email attachment model of embedding a copy of a file into the message ensures that data will be duplicated. This duplicity grows exponentially when coupled with the common end-user behavior of copying (Cc, Bcc) and forwarding messages to others, even to those who will never open the embedded file. Taking available data, we can estimate the number of files each business worker sends and receives per year at 17,125 [7]. For lack of data, this calculation assumes that an email with an attachment carries on one file, which clearly underestimates the total number of files duplicated in email. In a previous article, we calculated that around 6,000 of these attachments are completely ignored. These numbers represent only the amount of files represented by the user’s inbox, not accounting for copies in redundant servers and archival systems, which further multiply the quantity. Furthermore, those attachments that get read are often saved locally, thereby creating yet another copy in the recipient’s device. Finally, and also not accounted for in these statistics, is the additional duplication resulting from the recipient’s possession of multiple devices (e.g., laptop, mobile, etc.) — each which can receive its own copies. In an enterprise with one redundant email server and one email archive, we can comfortably assume that, at the minimum, each user represents more than 50,000 files per year.

Attachments are copied every time email is stored & attachments are accessed. File links only copy when files are accessed.

With email attachments representing, very conservatively, around 17,000 files per worker per year, we conclude that in a 1,000 person company a staggering 17 million file copies are created every year as a result of email’s file-sharing methodology — or about 50 million file copies given email redundancy and archives, but not accounting for the copies used by recipients outside of the email message. Email is data sprawl on an epic scale.

As a result of this explosive generation of duplicated content, most of it useless and all inefficient [3], our businesses, governments, schools and every organization that relies on email as their primary communication platform have to deal with an insurmountable challenge for information security, governance, and infrastructure cost.

On average, around 55,000 file duplicates created per business user per year as a result of email attachments

Security, Governance?

The data sprawl problem wouldn’t be as bad if the files sent through email were encrypted or tracked. Unless the user takes extra encryption steps, email attachments are completely unsecure and provide no tracking of their whereabouts or who is looking at them. Each of those 17,000 files per user per year is readily accessible to whoever has a copy of the email, and that includes the hackers who breach the email system, backup servers, a misplaced device, PST file, etc. — whether in your organization or any organization you communicate with. Furthermore, given that email attachments provide no possibility of revocation after being sent when systems are breached, valuable corporate content sent years before is still there.

Like smoke from a diesel truck, information leakage, and its resulting liability, is the natural byproduct of our emails

The anonymity of attachments also facilitates their usage as a means of cyber-attack by bad actors. Commonly classified as the primary vector of attack, email and its attachments provide an ideal delivery mechanism by which to bypass company defenses and get malicious code executed on the user’s device, well behind the company’s firewalls [4].

The sheer volume, distribution, and lack of inherent security of email makes controlling our data an impossible task

Companies deploy millions of dollars in data loss prevention and data security solutions, from requiring multi-factor authentication, installing network border systems, to providing regular employee information security training [5]. In addition, regulations require responsible custodianship of data and penalize heavily for infraction [6]. But how can the CIO, even the savviest armed with the biggest budget, contain a fundamental architectural flaw inherent in the Internet’s oldest and most widespread technology — an architecture that replicates, with viral efficiency, unprotected, untraceable copies of corporate content inside and outside of the organization, across multiple devices and systems!?


Actually, there is a solution, and there is a good chance you’ve used it. Partly in response to email’s limitations, the rapid growth of cloud content storage platforms (e.g., Box, Microsoft OneDrive, Google Drive, Egnyte) provides a new model of file exchange. The cloud storage model does not send a copy of the file, rather a link to the file stored by the sender. This model is the direct opposite of the email attachment model. The benefits of the cloud storage model are multiple. The epic data sprawl caused by email attachments is reduced by more than 75%. Now files are only distributed to those who are interested. Eliminating the aforementioned 6,000 files per user per year that go unread. If we include the elimination of copies sent, but never read, on different devices, backups and redundant email systems within and outside of the organization, we can estimate a reduction of greater than 90% in file sprawl. With the drastic reduction in distributed files, there is a concomitant reduction in network bandwidth and storage required to transmit and save unused files, plus an additional bonus reduction of 37% in file sizes from avoiding email’s archaic file encoding mechanism [3]. Just accounting for attachments in users’ mailboxes, available statistics attribute around 8Gb of files per business user per year — or around 7.8Tb for a 1,000 person company per year required by the email server, another 7.8Tb for any redundant servers, more for archival systems and more for attachments saved to local devices [1]. The use of cloud storage share links eliminates all the storage of attachments not viewed by recipients or copied to auxiliary email infrastructure.

Estimates put a year’s worth of email attachments for a 1,000 person company at 7.8TB or 20Tb with archives.

With regards to security, cloud storage links offer significant benefits for the sender and recipient. For the sender, access to files can be controlled, requiring authentication, allowing only viewing (no download), and expiring even after delivery of the message. For the recipient, cloud links provide a safe preview of file content away from their local device, eliminating the dangers of malicious code execution. They also provide origination by pointing back with a secure URL to an authenticated cloud storage account that has an owner, complete with a detailed audit trail, and security controls over how the file got there. In the event of a malicious file deployed from a cloud storage account, internal IT can more easily contend with a single URL than a massively distributed email message. Finally, file delivery is encrypted end to end — of interest to both parties.

Impact of Email Attachments — there is little to celebrate in email’s attachment file-sharing model

Call to Action

Like the oft-cited definition of insanity, businesses continue to dedicate significant time, money, and energy into problems originating from insurmountable challenges either directly caused or exacerbated by email. As the news cycle of email caused breaches proves to us on a daily basis, these efforts are of little avail. Until organizations recognize and resolve the root cause, they will continue their Sisyphean task, forever looking for and spending on the next technological miracle bandaid — be it Artificial Intelligence, Machine Learning, Social Graph, or another buzzword.

The reality is we are up against the fundamental architectural flaws of technology created decades ago and still in ubiquitous use today. These fundamental flaws have long been recognized. The seminal Internet Mail 2000 proposal posted by cryptologist and email technologist, Daniel J. Bernstein, outlined a new architecture whereby email would no longer be an analog to the postal service, i.e., where content is sent to the recipients, rather the recipients retrieve the content stored by the sender — a far more efficient design, simultaneously resolving email’s problems of spam, bounces, content proliferation, and false origination. However, redesigning the Internet’s most widespread communication infrastructure is a challenge. But we have entered a new moment, a moment where cloud content storage has become commonplace in the business world. With secure storage links, we correct the aforementioned flaws, bringing to fruition the known solution. Investments in user adoption training, increase in email client features for generating file links and services, like MXHero, that automate the creation of file links are far less costly than the perpetual, ineffective, and expensive game of whack-a-mole in common practice by organizations globally.

The reality is we are up against the fundamental architectural flaws of a technology created decades ago and still in ubiquitous use today.

Collectively we have a problem that is costing ourselves, businesses, and society billions and billions of dollars. We know the solution. For most organizations, cloud storage is readily accessible. It is time to prioritize and make it happen, namely, replace email attachments with shared links. We have so much more important things to do with our resources than to be trying to patch the collateral effects of a long-obsolete technology. Let us develop vaccines, solve global warming, educate our young. So please, for you, for me, for our world, no more email attachments.

When duplication is a good thing ;)

Originally posted to Medium


  1. Radicati

  2. CyberCrime Magazine

  3. Cisco

  4. Wikipedia

  5. Gartner

  6. SecurityToday

  7. Methodology & Calculations (spreadsheet download)


Products & Papers

Data Sheets



100 Pine St., Suite 1250
San Francisco, CA 94111