When looking at enterprise security, we commonly refer to and consider firewalls, Intrusion Prevention Systems (IPS), Virtual Private Networks (VPN), encryption and authentication. When we think of securing our data, we think of securing critical servers and databases. Rarely do we think of printers. Billions of dollars are spent worldwide on security each year, but how much did your organization spend on securing their printers this last 12 months? If you answered zero, you would be in the vast majority.
Printers have come a long brother printer driver way since their widespread adoption in the late 1970’s and early 1980’s. Back in the day, each printer was connected to an individual system and could only process a single print job at a time. Today, printers have matured into multi-functional devices that bare little resemblance to their distant origins. Printers in the 21st century perform dozens of tasks including, but not limited to, printing, scanning, photocopying, faxing and even emailing documents. What most users, and even system, network and security administrators do not realize is what really goes on inside a printer and what functionality they truly have. Most users still think of the printers of 30 years ago; unintelligent devices that only possess the ability to print documents. This view is far removed from the truth.
When discussing printers in this article, we are not only talking about the behemoths you see in most large enterprises, but also your low-end multifunctional printers you now find common in regular households. Rare is it to find a printer, no matter how small, that only performs the single task of printing. Most, at a very minimum, provide faxing or scanning and with these come increased memory requirements. Scanning a full document in preparation to print, scanning a document to be saved as a PDF or similar file, or scanning a document to allow faxing all require the ability to buffer the data within the device. A buffer is basically a region of memory that allows the storing of temporary data. Printers use this buffer to store a digital version of the document you are printing, scanning or faxing. Depending on the device, this buffer can range from a small piece of Random Access Memory (RAM) to a Hard Disk Drive like the type found in your desktop or laptop computer. In larger enterprise printers, this buffer is not the only memory store found within the printer. A larger, non-volatile memory area is provided to store semi-permanent or permanent information. For example, some printers allow scanning of a document and saving it within the printer as a PDF. The user may then connect to the printer as if it were a network drive, or via a web page, and download their document.
So where are we going with all this? The leakage or theft of sensitive and confidential corporate information. Large enterprises may have developed and implemented data retention and destruction policies but rarely do these include, or even mention, printers. Companies look at hardcopies of documents, CD’s, DVD’s and workstation, laptop and server hard drives when developing their data destruction policies. While it is clear they identify hard drives as a source of sensitive information, rarely do they consider the hard drives contained within their printers, if they even know of their existence. Printers are also commonly overlooked when security policies, procedures and guidelines are developed and implemented. Little time, if any, is spent looking at printer security or the implications of not securing the corporate printers. All the more disturbing this becomes when you contemplate the common types of documents that pass through printers in a corporate environment. Depending on the industry or the department within the organization, documents can vary from sensitive financial records, personal customer data or detailed network diagrams, to name a few.
To understand how sensitive data is leaked via a simple printer to the outside world, it requires an understanding of the corporate environment, security controls within that environment, and the general flow of information between users, printers and file systems that house restricted data.
In the ideal, secure corporate environment, a user has restricted access to files that pertain to his or her job function. The files reside on a secure server within the corporate network and are protected by strong access control policies requiring a user to authenticate before being allowed access to files. In our example, a user requires a sensitive financial document for a meeting he is about to attend. The user authenticates to the server, access to the file is authorized by the access control policies set on the file and the user opens the file in Microsoft Word. He clicks on the print icon and sends the document as a print job to his nearest printer. With this simple act, we have taken a secure document that very limited users have access to, and have created two copies that are no longer protected by any form of access control. The first is the obvious; the paper copy our user requires for their meeting. The second is a copy housed in the buffer on the printer. In the ideal world, our user will keep the printed copy safe at all times and follow the organization’s data destruction policy and destroy the copy of the document when they no longer require it. As for the virtual copy created on the printer, the user has no real control over this, nor probably knows it even exists. If we are lucky, the document is overwritten when the next print job comes through, but this is very dependent on the brand and model of printer and how the printer was initially set up by the administrator.
Slightly different to the straight printing of documents, scanning of documents or receiving faxes on a multifunctional printer writes documents to non-volatile areas of memory, usually a hard disk drive. If documents are not manually removed, they will remain there indefinitely, often long forgotten by the original user that scanned the document or received the fax.
In either of these scenarios, improper disposal of a decommissioned printer could have catastrophic consequences for a company. Leased printers may be returned to the leasing company for resale. Purchased printers are discarded in the trash or sold at auction or online via auction sites such as eBay. Either way, countless sensitive documents could pass into the hands of nefarious individuals. While the leaking of some documents could financially affect organizations, leaking personal information pertaining to hundreds or thousands of customers or clients could have reputation ramifications that could destroy a company.
Most organizations do not realize the full potential of their printers or the functionality they have available. While much functionality is non-security related, these functions have considerable impact on the security of the data within an organization and need to be understood and addressed. These include, but are not limited to:
1. The ability to copy files to Windows or Unix SMB file servers
2. The ability to email scanned files to a user
3. Functionality that allows printers to receive faxes and then forward the fax onto predefined users via multiple methods, such as email or as another fax, and
4. The ability to store files which have been scanned, printed, emailed or uploaded locally on the printer
While the previous data leakage scenarios have been accidental in nature, data remaining on printers could be the target of an educated attacker, one that understands the value of data residing on printers and who has the ability to compromise that data. While organizations invest hundreds of thousands of dollars to secure their network, dividing networks and systems into zones of trust with firewalls, Intrusion Prevention Systems and other network access control points, have they rarely considered where printers are logically placed within the network. In most cases, they are located amongst the users, or in some organizations, even on the server networks. Some organizations do not even have zones of trust and the printers exist amongst users, servers and even Internet accessible systems. In the worst case scenarios, the printers may even be Internet accessible themselves. Printers are not seen as critical devices, and as such, are not secured in their own zone of trust where access to management interfaces is not accessible except to trusted printer administrators. By limiting access to these interfaces, compromise of the data housed on these printers becomes exceedingly difficult.
While most printers have the capability to authenticate both printer administrators or normal printer users, the majority of the time, this functionality is disabled or left in its default state; disabled. Five minutes on Google and an attacker will be able to find the default password to almost any printer. Once administrator access is gained to a printer, it takes little time and even less ability to make changes to settings that could be catastrophic to an organization. While it would be little but annoying to find yourself locked out of your printer, or the interface changed to another language so no-one could control the printer, if the attacker was to redirect your printing or copy documents to a location outside the internal network, depending on the contents of the file, it could be the ruin of an organization.
So how does an organization protect itself against attacks against printers and leakage of sensitive data?
A few simple steps:
1. Disable unnecessary functionality. If any function within the printer is not required within your business, disable it. The less services or functions a printer has running, the less avenues of attack or leakage the printer has.
2. Add printers to your data retention and disposal policies. Make sure all memory inside printers is disposed of via secure destruction or secure wiping when printers are decommissioned.
3. Ensure data is overwritten immediately after printing. This requires the printer in use to support this functionality, but if your data is highly sensitive, this should be a priority when looking at new printers.
4. Print from memory rather than hard disk drive if available.
5. Use the secure printing option, if available, so printouts do not start before you reach the printer and enter your password. How often have you hit print, walked to the printer and your printout is no-where to be seen, only to turn up lying on a table days or even weeks later?
6. Examine where printers are logically located within the network. Printer management interfaces should be restricted and only accessible from defined management IP’s. Ensure printers are never accessible from the Internet. Assess whether some or all printers should be located within their own zone of trust.
7. Use the inbuilt security within the printer to restrict who has access, what access they have and where they may access from.
Securing printers should be an integral part of securing your data. Security policies should exist that address the risks and define how printers should be secured. Develop printer security guidelines and procedures for implementation of new printers and follow these standards to ensure all printers are secured and do not become a high risk to your organization. By securing your printers, you are contributing to your overall layered security model and protecting your organization’s critical data along with its reputation.