The thousand-fold increase in crypto-malware highlights a profound change in the cyber-landscape: Previously, an attacker seeking to steal intellectual property, personal identifiable information or payment card information would need to successfully breach and persist on one or more endpoints, carefully research the network, stealthily exfiltrate data, and finally process it in order to sell it on the dark web – a lot of effort for an uncertain payout. But crypto-malware is a clear signal that hackers have changed the game.
With crypto-malware – ransomware that encrypts files until a ransom is paid – every compromised device, whether company or personally owned, can be quickly monetized. If money isn’t the goal, attackers can use it to cripple a target for political or military advantage because it’s quick, precise and lethal, and much simpler and more effective than a messy kinetic weapon. For organizations whose missions depend on availability of computer systems, including hospitals, law enforcement and military targets, this new form of attack is a nightmare.
A crypto-malware attacker avoids the risk of post-breach detection or interrupted exfiltration by leaving data in place, but encrypting it without being detected. The attacker also maximizes fear and impact in a shocking way: via the ransom notice. That is, a personalized victim ransom page usually containing the initial ransom amount, instructions for how to purchase Bitcoins, and a countdown clock, which adds pressure to the victim by letting them know how much time they have to pay up before the ransom doubles or the data is deleted.
The attacker doesn’t even have to decide what data is valuable. Encrypting all data forces the victim to decide. Being both fearful and unsure, the victim is very likely to pay the ransom. Likewise, rather than having to sell stolen data at a discounted black-market price, encrypting it in place allows the attacker to directly demand top dollar from the victim, who values the data most.
Most importantly, crypto-malware breaches occur at machine speed, meaning there is no need for a remote human attacker to carefully dig deeper searching for valuable data. As soon as the endpoint is compromised the attack inflicts maximum damage. And since most such malware is undetectable by design, legacy AV suites offer scant assurance that you will be protected. Enterprise security teams have no opportunity to detect and respond to a breach as they do with traditional attacks.
Lest you think that crypto-malware is consumer focused, or that paying to decrypt your files is a simple way out, think again:
● Attacks have rapidly evolved to incorporate traditional breach techniques. Enterprise variants can propagate through the network to other devices and encrypt file shares and cloud storage to maximize impact.
● Following the classic approach of extortionists, attackers can charge different amounts to decrypt different parts of your data, or demand regular payments to keep data from being re-encrypted.
● It won’t be long before encrypted data is exfiltrated so the bad guys get to keep a copy too.
“Sorry, you’re going to be pwned”
Sadly, the security industry was already failing to protect its customers from traditional manual adversaries before attackers realized the benefits of machine-speed breaches. Vendors continue to peddle “maybe” technologies, like “next-gen antivirus” (NG-AV), that try to “detect to protect.” They tout their unbelievable ability to detect yesterday’s attacks – when in fact 99 percent of today’s malware morphs into new, undetectable variants in under a minute, according to the latest Verizon Data Breach Investigations Report. And they fail to protect against threats at the time of infection, instead offering remediation instead of prevention. Once the damage is done it’s too late.
Security vendors dodge responsibility for their failures, glibly encouraging their customers to continually look for signs of a breach they missed. But if you try to secure your organization assuming you will be breached by an adversary who operates on a human timescale, whose stealthy theft of data you must detect post-breach, you will undoubtedly be devastated by a machine-speed attack that cripples your organization before the first alert, then drains your bank account to “help” you back on your feet.
Protection at Machine-Speed
Today’s NG-AV tries to detect attacks and protect each endpoint individually, using signatures and heuristics updated by vendors on a human timescale. With absolute certainty this approach will fail, giving an attacker the foothold he needs to breach the enterprise at machine-speed.
To protect the entire enterprise at machine speed, you cannot rely on detection. The only solution is to protect “by design” -- architecting your environment to be resilient to attack. For a start, you ought to segment your network according to privilege or “need to know:” PCs that only access web applications need never be fully trusted on the enterprise network. Instead isolate them on their own VLAN or network subnet, and make users jump through authentication “gates” to get to high value applications or back-end services.
Don’t give users access to file shares that are not necessary, and keep database access limited to database applications. This concept of micro-segmentation of the enterprise network is being promoted by vendors of private and public cloud infrastructure, and the concepts apply on end user PCs through micro-virtualization. Ensure that users that need administrative access to corporate infrastructure are only able to elevate their privileges on a separately managed, VDI-based backend that can only be accessed from the enterprise intranet. Prohibit privilege escalation without forcing the user to log in under a different identity.
Rigorous separation of duties, with enforcement using tools from virtualization -- VDI, micro-segmentation and micro-virtualization -- are fundamental to building an enterprise infrastructure that is inherently more resilient to machine timescale attacks. Rigorous infrastructure-based enforcement of the principle of least privilege is a fundamental requirement for a resilient enterprise infrastructure architecture.
It is time to move beyond a model where we bet the security of the enterprise on the security of a single endpoint and human-timescale detection tools. The only way to defeat machine-timescale attacks is to embrace virtualization-enforced isolation to help enterprises protect themselves by reducing the enterprise attack surface.
- Crypto Ransomware Officially Eclipses Screen-Blocker Ransomware
- Improving Attribution & Malware Identification With Machine Learning
- Machine Learning In Security: Good & Bad News About Signatures
- 3 Flavors of Machine Learning: Who, What & Where