Dark Reading is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them.Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Endpoint Security

9/14/2018
07:00 AM
Joe Stanganelli
Joe Stanganelli
Joe Stanganelli
50%
50%

Lock Up Your Laptops: Cold Boot Attacks Are Back

Researchers at F-Secure have developed a workaround to nullify the popular ten-year-old patch that was thought to have solved the problem of cold-boot attacks. Encryption keys and other sensitive data on millions of laptops could be affected.

Reports of the death of cold-boot attacks have been greatly exaggerated. Cybersecurity wonks have long considered this a mostly solved problem, but new techniques have reanimated the decade-old vulnerability making cold-boot attacks possible.

On September 13, at the SEC-T conference, security consultants at Scandinavian cybersecurity firm F-Secure delivered a talk demonstrating that millions of laptops -- PCs and Macs alike -- remain vulnerable to a simple disk-encryption penetration attack involving a cold reboot and a USB stick.

The technique, which requires physical access to the target machine (and thus why laptops, portable as they are, are particularly vulnerable), is an updated version of a "cold-boot attack" -- a vulnerability known, and thought to have been patched, for over ten years.

(Source: iStock)
(Source: iStock)

Origins of the cold-boot attack In 2008, researchers at Princeton University published findingsthat DRAM in most then-modern computers, when without power (even if removed from the motherboard), retained enough data for a long enough time period for attackers to extract and reconstruct AES, DES and RSA keys. Effectively, this meant that attackers could penetrate popular disk-encryption systems just by forcing a hard reboot -- i.e., rebooting a computer "cold" instead of following a "proper" shutdown process -- and promptly removing the remaining data in the RAM.

The same year, the Trusted Computing Group released a specification for mitigating these attacks. The specification illustrated a BIOS-initiated operation for overwriting RAM on the next platform reboot -- making the matter of data remanence a moot point. Computer manufacturers implemented the specification, and the cold-boot vulnerability -- absent specialized equipment -- was thus considered solved. Then came Olle Segerdahl and Pasi Saarinen -- the F-Secure security consultants who discovered a way to disable this overwrite operation.

How Segerdahl got his cold-boot attackSegerdahl and Saarinen found that an attacker could physically tamper with a computer's firmware settings to rewrite the non-volatile memory chip and disable memory overwriting. From there, an attacker can enable external-device booting and carry out the classic cold-boot attack by booting specially designed malware off of a USB stick.

"Because this attack works against the kind of laptops used by companies there's no reliable way for organizations to know their data is safe if a computer goes missing," F-Secure security consultant Segerdahl wrote in a statement released Thursday. "And since 99 percent of company laptops will contain things like access credentials for corporate networks, it gives attackers a consistent, reliable way to compromise corporate targets."

The researchers called out Microsoft BitLocker in particular (their presentation's title? "An ice-cold Boot to break BitLocker") -- pointing to specific comments by Microsoft employees on the likelihood of success of data-remanence attacks on BitLocker-defended PCs.

"[I]n cases where the system memory is fixed and non-removable, [memory remanence attacks] are not possible using published techniques," reads a cached version of a Microsoft article from one year ago on BitLocker countermeasures. "Even in cases where system memory can be removed and loaded into another device, attackers will find the attack vector extremely unreliable[.]"

Segerdahl and fellow F-Secure security consultant Pasi Saarinen helped Microsoft to draft an updated version of the article. They have also announced that they "will publish techniques that allow recovery of BitLocker encryption keys from RAM on most, if not all, currently available devices." They further assert that these same techniques are similarly effective against other, non-BitLocker platforms (although Apple reports that Macs containing its T2 chip have security measures designed to mitigate attacks using these new techniques).

Shelter against the cold boot attack F-Secure reports that it has shared its research with Microsoft, Intel and Apple accordingly -- and that it is working with those companies on a patch. Still, Segerdahl has indicated that he is pessimistic on the probability of the industry fixing this remaining vulnerability in the near future. Moreover, while he concedesthat even with these newly developed techniques, cold-boot attacks remain "not exactly easy to do," Segerdahl maintains that they are still simple enough such that highly attractive targets -- such as banks and other large enterprise organizations -- should be on guard against them by implementing the following basic security measures:

  • Configure devices to require the user's BitLocker (or other disk-encryption) PIN every time they restore or power up their devices.
  • Eschew "sleep mode;" instead, configure laptops to automatically go into hibernate mode, if not shut down, when not in use (because encryption keys remain stored in RAM during sleep mode, but are not stored in RAM when a machine hibernates or shuts down).
  • Implement and maintain both a reporting policy and an incident-response plan specific to missing devices.

"Hibernation [and] pre-boot authentication [are] the best protection against cold boot attacks," tweeted Segerdahl earlier this month. "No keys in memory to steal!"

Segerdahl and Saarinen are scheduled to again present their findings at Microsoft's BlueHat Security Conference later this month.

Related posts:

—Joe Stanganelli, principal of Beacon Hill Law, is a Boston-based attorney, corporate-communications and data-privacy consultant, writer, and speaker. Follow him on Twitter at @JoeStanganelli.

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
Edge-DRsplash-10-edge-articles
I Smell a RAT! New Cybersecurity Threats for the Crypto Industry
David Trepp, Partner, IT Assurance with accounting and advisory firm BPM LLP,  7/9/2021
News
Attacks on Kaseya Servers Led to Ransomware in Less Than 2 Hours
Robert Lemos, Contributing Writer,  7/7/2021
Commentary
It's in the Game (but It Shouldn't Be)
Tal Memran, Cybersecurity Expert, CYE,  7/9/2021
Register for Dark Reading Newsletters
White Papers
Video
Cartoon
Current Issue
How Enterprises Are Assessing Cybersecurity Risk in Today's Environment
The adoption of cloud services spurred by the COVID-19 pandemic has resulted in pressure on cyber-risk professionals to focus on vulnerabilities and new exposures that stem from pandemic-driven changes. Many cybersecurity pros expect fundamental, long-term changes to their organization's computing and data security due to the shift to more remote work and accelerated cloud adoption. Download this report from Dark Reading to learn more about their challenges and concerns.
Flash Poll
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2022-23850
PUBLISHED: 2022-01-23
xhtml_translate_entity in xhtml.c in epub2txt (aka epub2txt2) through 2.02 allows a stack-based buffer overflow via a crafted EPUB document.
CVE-2021-4103
PUBLISHED: 2022-01-23
Cross-site Scripting (XSS) - Stored in GitHub repository vanessa219/vditor prior to 1.0.34.
CVE-2021-4172
PUBLISHED: 2022-01-22
Cross-site Scripting (XSS) - Stored in GitHub repository star7th/showdoc prior to 2.10.2.
CVE-2022-23807
PUBLISHED: 2022-01-22
An issue was discovered in phpMyAdmin 4.9 before 4.9.8 and 5.1 before 5.1.2. A valid user who is already authenticated to phpMyAdmin can manipulate their account to bypass two-factor authentication for future login instances.
CVE-2022-23808
PUBLISHED: 2022-01-22
An issue was discovered in phpMyAdmin 5.1 before 5.1.2. An attacker can inject malicious code into aspects of the setup script, which can allow XSS or HTML injection.