Simply put: this was a banner year for mega data breaches; according to Gemalto’s Breach Level Index, over 900 data breaches occurred during the first half of 2017, which compromised 1.9 billion records. More data was stolen in the first six months of 2017 than the entirety of 2016. Unfortunately, the massive exfiltration of data is a critical symptom of weak machine identity protection.
In the aftermath of major security incidents, experts often wonder how cybercriminals were able to exfiltrate large amounts of data while remaining undetected. Compromised machine identities allow attackers to use encrypted tunnels where traffic is only sporadically inspected - an approach that permits them to evade security controls. In fact, a recent study from A10 Networks found that 41 percent of cyberattacks use encryption to evade detection.
“Organizations increasingly rely on encrypted communication between rapidly changing networks of machines that are used for a wide range of critical-business functions,” said Nick Hunter, senior technical manager for Venafi. “To secure encrypted communications between machines, it’s vital that we protect each machine’s unique identity with at least the same rigor and precision we use to protect the online identities of humans. Unfortunately, most organizations don’t have the technology or intelligence necessary to do this, and because the number of machines on enterprise networks is exploding, this problem is rapidly getting worse.”
To highlight the role that compromised machine identities played in the data breaches revealed in 2017, we examined security incidents where large amounts of data were extracted without detection. Due to the massive scope and duration of these breaches, it’s likely machine identities played a pivotal role in these breaches.
- Uber Data Breach Impacts Over 50 Million Users and Drivers – According to Bloomberg the orchestrators of the attackaccessed a private coding site used by Uber engineers, and then used login credentials to access data from an Amazon Web Services account used by the ride-sharing company. Unfortunately, cyber attackers often abuse access to cloud services, such as Amazon AWS, to fly under the radar of other security controls.
- The Shadow Brokers Crack the NSA –Although the initial breach took place in 2016, the results of this incident were felt throughout 2017. For example, a data dump released in May revealed 300 megabytes of “cyber weapons" that allegedly originated from the NSA. In a distressing admission, Michael Morell, a former deputy and acting director of the CIA, said the NSA doesn’t know what else the leakers may have accessed, and, most concerning, whether the group is still actively stealing information. A key symptom of compromised machine identities is the inability to determine malicious activity inside encrypted tunnels.
- North Korea Steals War Plans from South Korean Ministry of Defense – According to South Korean Ministry of Defense department officials, nation state actors exfiltrated 235 gigabytes of information. The data may have included military documents and war plans, but roughly 80 percent of the stolen information had yet to be identified. Inability to decrypt and inspect encrypted traffic can allow attackers to exfiltrate large amounts of data while undetected.
- The Yahoo Breach Expands to 3 Billion Users – After the first reports of a breach at Yahoo!, Venafi Labs discovered that Yahoo! appeared to have weak machine identity management across many of its external web properties. Weak cryptographic security can allow attackers to steal intellectual property, elevate privileged access, eavesdrop on confidential communications and retain backdoor access to networks for extended periods of time.
“Effective machine identity management requires complete visibility and continuous assessment of all identities across the extended enterprise,” concluded Hunter. “Only comprehensive intelligence can drive the automated, coordinated actions that are necessary to proactively remediate machine identity weaknesses. Until more organizations have these capabilities, we will continue to see massive breaches, even among large organizations with major investments in security. The only way organizations can stem the rising tide of data loss is to automate machine identity management.”
Will we continue to see breaches that manipulate weak machine identities?