Vulnerabilities / Threats

7/28/2017
05:00 PM
Connect Directly
Twitter
LinkedIn
Google+
RSS
E-Mail

Lethal Dosage of Cybercrime: Hacking the IV Pump

At DEF CON, a researcher demonstrated how to attack a popular model of infusion pump used in major hospitals around the world.



DEF CON - Las Vegas - The dangers of medical device hacking can be severe, and potentially lethal, when conducted by cybercriminals with the means and motivation to cause physical harm.

Dan Regalado, principal security researcher for IoT firm Zingbox, demonstrated how an attacker could break into a medical infusion pump during a presentation at DEF CON's IoT Village. IV pumps are sensitive and pricey medical devices used to deliver medications, fluids, blood, and blood products to adult, pediatric, and neonatal patients.

Infusion pumps present a growing attack surface for hackers to break into at hospitals. The market is expected to hit $10.84 billion by 2021, he said, citing a "Market and Markets" forecast.

Why would someone hack an IV pump? There are several reasons, Regalado pointed out. If successful, an attacker could steal personally identifiable information (PII), hijack hospital devices and demand ransom, corrupt the device in a denial-of-service attack, or use the pump as an entryway into the broader corporate network. 

For his educational research, Regalado chose to break into the Alaris PC Unit and IV Pump module manufactured by Bectron. The pump is a market-leading brand used at several hospitals around the world, he said.

His presentation dug into details of the device's internal components and their vulnerabilities. 

Flash FX, the generic block device driver, links the system and hardware. It lets the OS access the internal flash memory as if it were a hard drive or RAM disk, and stores sensitive information like credentials for accessing network systems. The compact Flash card inside the pump, used to boot the ENEA OSE system, stores patient profiles and is easy to retrieve and alter.

Regalado demonstrated onstage how an attacker could bypass the image integrity check, and gain access to the restricted configuration by changing the machine's PIN number.

"You can modify the integrity of the system," he explained. "The infusion pump will do whatever you want."

From there, a hacker could conduct a man-in-the-middle attack by impersonating the access point and server to manipulate pumps. Each hospital can have up to 400-1,000 pumps, said Regalado, each of which could have its settings configured to administer dangerously low or high dosages to its patient.

An attacker could also use this level of access to reconfigure the pump's network properties and overwrite the internal flash with new wifi configurations, or execute commands from the internal shell to destroy critical files or collect sensitive data.

"We need to assume a physical attack will happen and work towards asset protection," said Regalado, adding that internal attacks are more common than external ones.

While this demonstration required physical access to the pump, he warned remote attacks would be next. Regalado also noted that the vendor was notified of the pump's vulnerabilities and committed to a 30-60 day policy to address them.

His presentation ties in with the broader growth of cybercrime on healthcare organizations. Major cyberattacks on the industry grew 63%. Researchers discovered an increase in medical device hijacking, which involves the use of backdoors in medical devices to load malware tools and exfiltrate intellectual property. 

The same year, a record-breaking 328 healthcare businesses reported data breaches. Experts say attackers are becoming more confident as breaches are increasingly publicized. It's becoming clearer that healthcare targets aren't as security-savvy as once believed, and the industry has consistently seen more breaches year after year.

Related Content:

Kelly Sheridan is the Staff Editor at Dark Reading, where she focuses on cybersecurity news and analysis. She is a business technology journalist who previously reported for InformationWeek, where she covered Microsoft, and Insurance & Technology, where she covered financial ... View Full Bio

Comment  | 
Print  | 
More Insights
Comments
Oldest First  |  Newest First  |  Threaded View
Christian Bryant
50%
50%
Christian Bryant,
User Rank: Ninja
7/31/2017 | 12:16:25 PM
Risk Versus Benefit Analysis
This is an excellent example of automation where the question of risk versus benefit really needs to be called out.  Do we need software in everything?  In terms of lives saved, are we really saving more lives through IV automation than without?  Or is money that is being saved?  Because in healthcare IT all that (should) matter is whether lives are being saved or health promoted.  If these IVs are automated due only to some perception money is being saved, think again.  Consider the millions needed to firewall and protect your IV pumps, to address intrusions and potential ransom demands.  I think every piece of automation that puts lives at risk should be re-examined and a determination made whether just returning to low-tech methods isn't more sensible in the long run.
More Than Half of Users Reuse Passwords
Curtis Franklin Jr., Senior Editor at Dark Reading,  5/24/2018
Is Threat Intelligence Garbage?
Chris McDaniels, Chief Information Security Officer of Mosaic451,  5/23/2018
Register for Dark Reading Newsletters
White Papers
Video
Cartoon Contest
Write a Caption, Win a Starbucks Card! Click Here
Latest Comment: This comment is waiting for review by our moderators.
Current Issue
Flash Poll
[Strategic Security Report] Navigating the Threat Intelligence Maze
[Strategic Security Report] Navigating the Threat Intelligence Maze
Most enterprises are using threat intel services, but many are still figuring out how to use the data they're collecting. In this Dark Reading survey we give you a look at what they're doing today - and where they hope to go.
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2018-11506
PUBLISHED: 2018-05-28
The sr_do_ioctl function in drivers/scsi/sr_ioctl.c in the Linux kernel through 4.16.12 allows local users to cause a denial of service (stack-based buffer overflow) or possibly have unspecified other impact because sense buffers have different sizes at the CDROM layer and the SCSI layer.
CVE-2018-11507
PUBLISHED: 2018-05-28
An issue was discovered in Free Lossless Image Format (FLIF) 0.3. An attacker can trigger a long loop in image_load_pnm in image/image-pnm.cpp.
CVE-2018-11505
PUBLISHED: 2018-05-26
The Werewolf Online application 0.8.8 for Android allows attackers to discover the Firebase token by reading logcat output.
CVE-2018-6409
PUBLISHED: 2018-05-26
An issue was discovered in Appnitro MachForm before 4.2.3. The module in charge of serving stored files gets the path from the database. Modifying the name of the file to serve on the corresponding ap_form table leads to a path traversal vulnerability via the download.php q parameter.
CVE-2018-6410
PUBLISHED: 2018-05-26
An issue was discovered in Appnitro MachForm before 4.2.3. There is a download.php SQL injection via the q parameter.