This article could start by reviewing the college that closed permanently after a ransomware attack or the large school district that suffered an incident during a recent holiday weekend. Or it could focus on how critical infrastructure—such as water treatment plants, pipelines, and meat processing plants—are increasingly under attack. We could even comment on the increase in ransom demands.
We are not going to do any of the above—at the beginning of this article or anywhere else. While there are important lessons to take from each of these situations, we will leave that analysis to others.
This article takes a different approach. It starts by quickly establishing the current context on ransomware before moving into an analysis of where current and best practices diverge. We’ll look at eight areas where many organizations remain susceptible to ransomware attacks, outline new and emerging solutions or approaches that can be used to bolster controls and protections, and offer a report card for self-assessment by organizations. Most report cards are additive (the “better” level also requires the “baseline” controls, and the “best” level requires the controls from all three levels) while two are based on maturity (low, medium, and high).
The first four areas focus on defending against ransomware attacks, the final three focus on recovering after an attack, and the fifth area does double duty for defending and recovering. By the end of this article, decision-makers and influencers charged with evaluating and selecting cybersecurity solutions should have a better idea of their organization’s readiness (or not) to counteract ransomware.
Ransomware in 2022: Snapshot
Are Attacks Increasing or Decreasing?
1. Strengthen Identity Methods
2. Act on Internal Threat Signals
3. Extend Threat Intelligence
4. Human Layer of Defense
5. Add Resiliency to Data Backup
6. Prepare to Recover After an Incident
7. Prevent Abuse of Exfiltrated Data
8. Develop Resilient Operating Processes
Ransomware remains a threat—and a growing one at that
Threat actors are evolving their toolkits and playbooks to make ransomware more devastating to victims. Ransomware-as-a-Service (RaaS) and partner-incrime models, along with greater supply chain specialization, increase the peril.
Metrics for ransomware attacks are both up and down
Some reports on ransomware show a threat with increasing frequency. Others show a downward trend. Sanctions imposed on Russia following its invasion of Ukraine appear to have neutralized the efficacy of Russian-based gangs.
Strengthen defenses against ransomware attacks
If 2022 is a “strategic pause” for ransomware attacks, organizations should increase security posture and decrease threat susceptibility for when ransomware gangs return to crime-as-usual.
Prepare to recover after a ransomware incident
While it is almost certain that every organization will face opportunistic attackers—and perhaps determined ones too—it is not certain that every organization will become a victim of a ransomware incident. But if that does happen, having a response protocol or recovery plan approved by the Board and ready to enact is invaluable. Lay a strong foundation now for recovery, if needed.
Ransomware in 2022: Snapshot
Ransomware attacks are associated with a common set of realities in 2022. In brief:
Multiple levels of extortion increase the likelihood of a massive payday
Threat actors seek financial gain from ransomware attacks. It is bad news for the threat actor if the victim organization does not pay the ransom demand. To increase the likelihood of receiving a payment, threat actors have innovated beyond malicious data encryption as the sole blackmail lever. Additional levers include threatened distributed denial of service (DDoS) attacks and data exfiltration through hard-to-detect mechanisms like DNS tunneling with threats to sell, publish, auction, or otherwise disclose the stolen data if the ransom is not paid. Victim organizations may be able to address the operational disruption of malicious data encryption through backups, but mitigating risks associated with DDoS and exfiltrated data is a different challenge entirely.
RaaS and partner-in-crime models expand the number of attackers, ransomware variants, and advanced threat methods
Threat actors are making ransomware toolkits easily available to wannabe threat actors, speeding up their time to crime. Today’s toolkits offer advanced evasions, exploits, and other techniques that were previously available only to nation-state actors or large cybercrime and ransomware gangs. When a new entrant threat actor lands a successful attack using a RaaS toolkit, any ransom payment is divided between the threat actors. RaaS has been implicated in increasing the number of ransomware attacks. The number of ransomware variants in the first half of 2022 doubled compared to the second half of 2021.
Ransomware supply chain models amplify specialization through division of labor
Ransomware gangs are embracing specialist skills in crafting ransomware attacks, such as Initial Access Brokers (IABs) who sell access to compromised networks or devices, and post-compromise negotiators fluent in the language of the victim organization to negotiate the ransom payment. Ransomware gangs seek the same benefits from specialization and division of labor that the world has seen in most other business endeavors, such as manufacturing.
The most common entry points for ransomware are phishing (for credentials), open network ports and virtual private networks (VPNs), and vulnerabilities in applications
Distracted employees clicking malicious links, misconfiguring a device, or forgetting to apply a critical security patch continue to be the leading causes of most breaches. Email is still the top attack vector so businesses must evolve their thinking beyond email perimeter defense and user training. Insufficient protections against these attack entry points is like painting a target on the front door.
Paying the ransom is expedient, but often ineffectual
Many organizations pay the ransom demand to gain the decryption key, stop the publication of exfiltrated data, or prevent a DDoS attack—despite low success rates. Cyber insurance cover is a leading source of funding for paying ransoms. However, decryption keys don’t always work to restore encrypted data and cyber criminals still leak data. Paying the ransom to recover from an attack has four significant weaknesses. First, it doesn’t address the systemic shortcomings that allowed the attack to succeed. Second, it funds additional cyber crime. Third, future cybersecurity insurance premiums will be higher— assuming coverage even remains available to the organization given their changed risk calculus. Finally, it often leads to additional attacks or subsequent ransom demands for exfiltrated data because threat actors know the organization has a proclivity to pay up.
Just one unaddressed entry point is enough for an attack to gain a foothold
Ransomware attacks start with one successful phishing email, access to a single password to a legacy VPN that isn’t using multi-factor authentication (MFA), or abuse of a newly identified or unpatched software vulnerability. The problem has been exacerbated due to visibility challenges stemming from the transition to a more hybrid workforce over the past couple of years. Hoping that threat actors will not trick employees, find infrequently used systems, or attack irregularly patched applications has repeatedly proven shortsighted.
Are Attacks Increasing or Decreasing?
Several reports suggest ransomware attacks are increasing in 2022:
August and September 2022 were two of the three worst months for ransomware incidents since January 2020
BlackFog tracked 39 ransomware reported incidents during August and 33 in September, two of the three highest numbers recorded since January 2020. In July, BlackFog concluded that many incidents were going unreported, since public notification of incidents were decreasing while attacks were increasing.
Increased activity among RaaS threat actors during 1H 2022 vs. 1H 2021
Trend Micro’s 2022 Midyear Cybersecurity Report shows significantly higher ransomware detections from several threat groups, including LockBit (5x higher), Conti (almost 2x higher), and BlackCat (a new group at the end of 2021, but already with 75% the number of detections as LockBit). Competition between RaaS threat actors will spur the development of improved toolkits.
Living through the ‘golden era’ of ransomware
European authorities said the world was facing the golden era of ransomware, based on an analysis of incidents from April 2020 to July 2021 (which increased by 150%). They warned that ransomware was likely to get worse before the era ends, calling it the “prime threat” facing organizations as cyber criminals follow the money to profitable criminal activity.
Other reports indicate attacks are trending downward. For instance:
23% decline in ransomware attacks in 1H 2022
The mid-year update to the 2022 SonicWall Cyber Threat Report shows a 23% decline in the number of ransomware attacks from January 2022 to June 2022 compared to the same time period in 2021 (236.1 million in 1H 2022 vs. 304.7 million in 1H 2021). While that is good news, SonicWall offers the context that the volume of ransomware in 1H 2022 is already higher than the full-year totals for 2017, 2018, and 2019, and is 77% of the way to reaching 2020’s full-year total.
7% decline in 2Q 2022 vs. 1Q 2022
The number of ransomware victims tracked per month decreased by 7% from Q1 (232 victims per month) to Q2 (216 attacks per month).
NSA on declining ransomware attacks
Russia’s invasion of Ukraine has disrupted operations of ransomware threat actors and gangs, many of which operate in Russia. The National Security Agency (NSA) in the United States says that financial sanctions make it more difficult for Russian gangs to capture ransom payments and purchase infrastructure to run new attacks. It is uncertain when the invasion will end, and it is uncertain to what degree and with what speed the sanctions will be lifted when it does. But the net current effect is a decline in the ability of the ransomware ecosystem to perpetuate new attacks.
One interpretation is that 2022 is a “strategic pause” in ransomware attacks. However, irrespective of whether attacks are increasing or decreasing, organizations should strengthen what is working and address what is not. When attacks are increasing—in frequency, sophistication, and damage potential— urgency is high for rapid mitigation. When attacks are decreasing or paused, there is an opportunity for a more thorough readiness review. Doing so increases security posture and decreases threat susceptibility.
In this section, we present a SWOT on ransomware in 2022.
Wider understanding of the ransomware threat
General awareness of the ransomware threat is much higher than five years ago, and many organizations have made good progress in mitigating outstanding vulnerabilities and weaknesses that were open to attack.
Higher systemic preparedness
Protections that address ransomware-specific threats create wider protections against other types of cyberattacks. For example, stronger forms of MFA reduce the likelihood of network breaches in a ransomware attack, as well as decreasing vulnerability to phishing attacks. Likewise, investments in zero trust, least privilege, and network segmentation have strengthened network security—and hardened defenses against ransomware.
Ransomware attacks are still evolving and successful
Ransomware attacks continue to turn into incidents, affecting organizations across all industries. New vulnerabilities and exploits amplify this challenge.
Preparedness against key entry points is lacking
While organizations have made some progress in strengthening protections against ransomware, many hold back from embracing stronger protections until after suffering a devastating incident and senior management feeling the gravity of the risks enumerated by security leaders. Almost all organizations promise to strengthen identity and embrace MFA in their press release after an incident.
Embrace the opportunity of the strategic pause
Use the current strategic pause in ransomware attacks to strengthen defenses, plug gaps in visibility, address vulnerabilities, and decrease the likelihood that a future attack will succeed.
Decrease relative attractiveness of attacking your organization
Protecting against ransomware requires pursuing two outcomes: decreasing the attractiveness of each organization as a target relative to all other organizations and decreasing profitability of ransomware as a criminal endeavor. The first outcome can be pursued directly; the second only indirectly.
Ransomware gangs hampered by sanctions and seeing fewer successful attacks for their efforts may inflict greater damage, escalate ransom demands, and impose retaliatory long-run effects on the few they do compromise. For example, they may pivot from malicious data encryption to permanently wiping data or destroying physical control systems. Further, once the invasion of Ukraine ends, Russian gangs are likely to significantly increase their attacks to regain profits lost during the conflict.
Increasing rates of data exfiltration
Virtually all ransomware incidents include data theft to increase the likelihood that the ransom demand will be paid. Threat actors use stolen data in serial ransom demands. Too few organizations consider protections against data exfiltration as a mandatory part of their cyber toolset.
1. Strengthen Identity Methods
Account credentials provide access to systems, applications, and devices—unlocking a set of access privileges depending on the user for whom the credential is intended. But a username and password are associated with an individual only indirectly; there is nothing that directly binds the credentials to an individual. Credentials can be compromised through credential stuffing attacks or by tricking an employee into giving up their credentials through a phishing email. This allows a threat actor to present themselves to systems, applications, and devices as the employee. Phishing is a frequent precursor to ransomware attacks, because ownership of an email account establishes a trusted identity that can be used to reset passwords to other systems, or phish other employees and vendors. Strengthening identity with MFA is almost always undertaken with urgency and a sense of mea culpa after a ransomware incident.
SOLUTIONS FOR STRENGTHENING IDENTITY
MFA requires multiple forms of authentication before access is granted. MFA links something the individual knows (the username and password) with something they have. The “have” part is provided through a separate device when needed. Basic forms of MFA rely on codes sent by SMS or email, but these have proven vulnerable to compromise. A threat actor can access these codes if they have compromised the email account or succeeded with a SIM-swapping scam. Stronger options are:
Shift to stronger forms of MFA
Stronger forms of the “have” part of MFA include one-time passcodes generated by Authenticator apps on mobile devices and FIDO-based hardware keys. While these forms are much stronger than codes sent by SMS or email, they have two shortcomings. First, unless biometric checks are also required (such as a facial scan on a device), mere possession of the mobile device or hardware key by a threat actor still enables identity theft. Second, many phishing kits can capture and leverage codes from authenticator apps.
Use biometric authentication to tie identity with a specific person
While MFA relies on a “have” for the second factor, biometric authentication relies on an “are” factor. Assuming non-commodity solutions are used to create and confirm biometric indicators, voice prints, fingerprints, and face scans enable unique identification of an individual. Embracing stronger forms of identity decreases the opportunistic threat of compromised credentials by significantly increasing the difficulty of abusing stolen credentials.
Embrace risk-based authentication
Authentication based on the strongest forms of identity will sometimes not be enough. Other risk factors beyond the identity—the type of device used, the network connection, when access occurs, and the geographical location of the access request—can signal greater need for caution in granting access at all or in varying gradients. Solutions that unlock visibility into these signals enable the use of additional authentication and verification checks when risk is high.
REPORT CARD FOR STRENGTHENING IDENTITY (MATURITY)
- Low: Username and password with basic forms of MFA
- Medium: Stronger forms of MFA such as Authenticator apps and hardware keys
- High: Biometric authentication plus risk-based authentication
2. Act on Internal Threat Signals
Stopping ransomware from establishing a foothold inside the organization’s systems is an essential cybersecurity strategy, but it cannot end there. If ransomware does slip through defensive armaments, neutralize it before it causes damage. Having the visibility to continuously identify and respond to internal threat signals is part of a comprehensive defensive strategy.
SOLUTIONS FOR INTERNAL THREAT SIGNALS
Solutions that enable detection and mitigation of internal threat signals include:
Ensure visibility of internal attack surface
Endpoint detection and response solutions (along with newer variants) profile applications running across the IT ecosystem to provide visibility on what employees are using and how widely. This enables rapid detection of vulnerable applications, versions, and processes considering new threat data. Such solutions also support detection of lateral movement before ransomware is detonated. After gaining an initial foothold on one device through compromised credentials or exploiting a vulnerable application, threat actors attempt to install malware and hacking tools as widely as possible. The more widely predetonation malware is installed, the more crippling the attack when it is denotated. While this process of lateral movement is stealthy it is not invisible, hence if detected, it can be mitigated before the malware is detonated.
Automate detection of and response to mailbox threats
Attackers have mastered the ability to deliver threats to employee inboxes and trick distracted employees into clicking a malicious link. Managing this problem of evasive and sophisticated email attacks requires coordinated optimization across detection methods, user participation, automation, and security analysts. Achieving a mature approach to targeted email attacks requires automating processes, shifting from prevention to incident response, modernizing user training to apply it to current attacks in real-time, and leveraging managed services to reduce alert fatigue and email security configuration errors by security analysts.
Monitor baseline deviations to detect insider threats
At least one ransomware gang is actively recruiting employees to become turncoats and provide surreptitious access to the IT systems of their organization as the entry point for a ransomware attack. User and Entity Behaviour Analytics (UEBA) solutions create a baseline of normal behavior. These solutions track access patterns for employees—individually and in context of the group/department they belong to—and can throw off early warning signals when an employee’s normal baseline is subject to sudden, strange, or unexplained variations.
Interrupt communication with command-and-control infrastructure and data exfiltration activity as it is happening
Prior to detonation, ransomware needs to communicate with its commandand-control infrastructure. Anti-Data Exfiltration (ADX) solutions identify outbound communication attempts from pre-detonation ransomware to this control infrastructure. Once identified, ADX solutions enable communication attempts to be disrupted, and by implication, the receipt of any returning commands to be eliminated. Such solutions should monitor outbound traffic from endpoints to identify abnormal behavior in traffic patterns and signals of compromise that are reflective of malware, ransomware, and other suspicious processes. These include endpoint processes that were not designed to initiate network traffic, communication attempts with ransomware hotspot locations (such as Russia, North Korea, and China), and reliance on dark web protocols.
Stop malicious data encryption as it is happening
Solutions that detect the detonation of malicious data encryption can halt ransomware in its tracks. As soon as malicious encryption is detected, such solutions stop it and restore encrypted files to their pre-encrypted state, as well as triggering high urgency alerts about the detected ransomware activity.
Leverage network data for security visibility and control
Network visibility helps defenders who find themselves caught between the evolution of ransomware evasion techniques and the explosion of devices— such as bring-your-own-device strategies (BYOD), the Internet of Things (IoT), industrial control systems (ICS), and operational technology (OT). DNS-level defenses counter spoofing, lookalike domain names, DNS tunneling, and other evasion techniques both on and off-network. Some attacks simply use DNS since it is mostly unmonitored by secure web gateways (SWG), next-generation firewalls (NGFW), and other defenses. Tools like IP address management (IPAM) can provide responders with on-demand access to critical device details—such as platform and operating system details—to help prioritize and frame effective responses.
There are too many threats across too many vectors to delegate primary responsibility for acting on internal threat signals to cybersecurity staff, whether internally or at a managed services provider. No organization can afford the time or money to hire enough staff to manually address all ransomware threats across all elements of their IT ecosystem—networking components, servers, endpoints, and cloud services. The primary workflows for detection, neutralization, mitigation, and incident response across the breadth of the IT ecosystem must rely first on advanced automation. This has two implications:
- Cybersecurity staff complement these workflows and make decisions or judgment calls when needed in automated processes.
- Solutions that capture internal threat signals but require security analysts to take action are less valuable than solutions closer to the fully automated end of the continuum.
The indicators in the report card are additive since they work together to assure overall security.
REPORT CARD FOR INTERNAL THREAT SIGNALS (ADDITIVE)
- Baseline: Solutions that automatically interrupt early-stage communication with command-and-control servers, stop data exfiltration and malicious data encryption activity, prevent the delivery of threats to employees through email, and automatically mitigate vulnerabilities detected across endpoints, servers, cloud services, and more
- Better: Detect and respond to inbox and insider threats or compromise
- Best: Visibility into endpoints, servers, cloud services, and more for human-driven analysis of potential weaknesses, vulnerabilities, and egress points
3. Extend Threat Intelligence
Threat intelligence collates and synthesizes details of current threat activity experienced by a cohort of organizations. Vendors, governments, open-source groups, and others offer threat intelligence resources to provide early warning signals on current attacks, pending targets, and vulnerabilities that are actively exploited or otherwise high risk. Threat intelligence is based on the principle that if attackers are doing it, planning it, or know about it, organizations must too. It is imprudent for organizations to have less visibility than attackers.
SOLUTIONS FOR EXTENDED THREAT INTELLIGENCE
Solutions to explore for extending threat intelligence:
Hunt for threats across your ecosystem
Proactive threat hunting is the practice of looking for cyber threats that have gained an early foothold in your ecosystem. Threat hunting correlates parameters from multiple sources to identify the origin, type, and frequency of an attack. Leveraging aggregated data signals from multiple organizations makes it possible to obtain detailed insight into each identified threat, so that organizations stay ahead of cybercriminals as the threat landscape evolves. Threat hunting tools and services enable organizations to identify false positives, investigate threat origins, and provide peer-based risk analysis.
Monitor the external attack surface
Threat actors seek to discover and exploit weaknesses in the ecosystem, systems, and applications they can access. Government agencies offer complimentary external attack surface monitoring to certain organizations, and insurance providers are using these same tools to assess risk when evaluating coverage levels and setting premiums. Since it is becoming an accepted part of assuring and assessing defensive posture by government agencies and insurance providers, it is the type of capability that organizations should be using themselves. Organizations that embrace solutions and/or managed security services that offer proactive and continuous external attack surface monitoring are more likely to stay ahead of threat actors in identifying weaknesses to mitigate and vulnerabilities to address. When threat actors can see what organizations cannot, attackers have a strategic advantage.
Track emerging threat indicators on dark web forums
Chatter on dark web forums offers early warning of potential attacks. Threat actors, particularly IABs, offer to sell access to networks or devices at newly compromised organizations. While company names are not disclosed in the sale notice, attributes of the compromised organization are listed. Some threat intelligence services are finding a modicum of success in correlating attributes with a given organization.
REPORT CARD FOR EXTENDED THREAT INTELLIGENCE (ADDITIVE)
- Baseline: Threat intelligence services
- Better: Threat hunting for deep intelligence and prevention
- Best: External attack surface monitoring and dark web forum monitoring
4. Human Layer of Defense
People play a key role in protecting the organization from ransomware attacks by having the ability to detect suspicious phishing messages and other social engineering attempts, for example when someone from the “IT help desk” calls and asks for the employee’s password or MFA code. Another part of this role is actively embracing the security protections required by the IT or security team to safeguard the organization, such as use of biometric hardware tokens. While people must never be treated as the only means of defense, they play an essential role as one of the defensive layers.
SOLUTIONS FOR CREATING A HUMAN LAYER OF DEFENSE
Creating a human layer of defense relies on the following approaches:
Offer better security awareness training
Forewarned is forearmed, and enlisting employees, managers, and executives in the war against ransomware is a must. Many organizations continue to offer security awareness training too infrequently and too rigidly to be helpful. Employees view such training as a check-box activity and remain disengaged—to the detriment of the organization. Static curriculum, exams, and fake phishing emails haven’t stopped users falling for mailbox threats. Better security awareness training is characterized by relevancy to the individual and group, teachable moments, and frequency of small learning chunks. It must also address signals of an in-progress attack, such as receiving unexpected MFA prompts.
Test the training and act on the quantified risk
Training on security and testing the efficacy of that training are two sides of the same coin. Platforms that offer training content should also include testing capabilities to evaluate which individuals and groups struggle to apply the concepts. Testing approaches include phishing simulation, employee susceptibility to USB thumb drives “dropped” in the company parking lot, and penetration testing of facilities and offices. Learnings from these tests quantify current risk and provide insight for subsequent training and education initiatives. Distraction is a risk, too, and one that security awareness training often ignores.
Close the loop with employees on reported threats
Part of security awareness training is establishing the organizational processes for employees to report suspicious emails, particularly suspected phishing emails that are often a precursor to credential compromise and ransomware gaining a foothold. But it is what happens after they click the button to report the email that really counts. If an employee hears nothing back from the IT or security group that receives reports of suspicious emails, what does that mean and what should they do? Ensuring the IT or security group is sufficiently resourced is essential. This requires sufficient personnel, as well as optimal tooling that automatically distinguishes malicious reported emails from the benign, remediates malicious messages across all email inboxes, and closes the loop with employees.
REPORT CARD FOR A HUMAN LAYER OF DEFENSE (ADDITIVE)
- Baseline: Relevant, context-aware security awareness training
- Better: Training efficacy is regularly tested and evaluated
- Best: Reported threats are mitigated promptly; training infused with current topics and is applied to real-time threats
5. Add Resiliency to Data Backup
A primary goal of ransomware attacks is monetary gain. Threat actors design the attack so that the victim organization pays the ransom to recover their data promptly. Backups have proven an effective countermeasure for organizations, since access to backups means systems, applications, and devices can be recovered without paying the ransom. For this reason, threat actors have extended their toolkits and playbooks to gain access to, delete, and corrupt backups by ensuring dormant ransomware is backed up, too. One study found that backup repositories were targeted in 94% of ransomware attacks and at least a portion of repositories were impacted in 68% of incidents. If ransomware gangs pivot to using data wiping malware rather than just data encryption malware, backups will become critical to recovery—as paying a ransom for the decryption key is no longer an option when the data has been wiped.
SOLUTIONS FOR RESILIENT DATA BACKUP
Many organizations have a backup strategy, but only a minority successfully recover their data after a ransomware attack. Organizations that want to avoid downtime, streamline recovery, and not cave to paying the ransom to get their data back should be looking at the following solutions:
Test that backups work and enable restoration
Only one third of organizations can recover 80% or more of their data after a ransomware incident. Betting the future of the organization on backups that don’t enable full recovery within a short period of time is more costly than doing it right in the first place. If backups are going to be part of an organization’s strategy against ransomware, they need to work, have been subjected to repeated testing, and support recovery dependencies between systems.
Use ransomware-aware intelligent backup and restoration procedures
Backups don’t help in recovering from a ransomware attack if backups are riddled with dormant ransomware ready to denote upon restoration. Restoring non-encrypted business files along with dormant ransomware from infected backups merely enables the ransomware attack cycle to start again. Backup solutions need the intelligence to differentiate between business files and ransomware exploits if a clean backup is to be created or restored. All data should be scanned for malicious code during backup procedures and checked again during data restoration processes.
Elevate controls over changing backup settings and deleting data
Stronger identity controls are needed for approval and confirmation of significant requests against backup repositories. These added controls thwart threat actors who manage to compromise an administrator’s account from maliciously changing backup and retention settings, or deleting backup data.
Adding resiliency to data backup does double duty for defending against and recovering from a ransomware attack. It is a defensive strategy because resilient data backups mitigate the threat of devastating consequences from a ransomware incident by safeguarding business data. It is also a recovery strategy, because if the organization is compromised through a ransomware incident, resilient backups provide a mechanism for restoring business data without paying the ransom for a decryption key.
The emphasis with resilient data backup is the backup of data in a resilient way, which is more than just the backup of data stored on servers in a resilient way. Data must be incorporated into backup processes wherever it is stored, so that the full complement of data required to operate business processes and meet archiving and compliance mandates are met. Designing backup processes so that only servers are included will by design exclude all data created on endpoints that is not synchronized to a server or cloud service.
Business data created by employees on laptops and mobile devices that is stored independently of a sanctioned business system will be ignored in backup processes if these endpoints are excluded. Such data can include notes from client meetings, early-stage product design ideas, and draft business reports. The shift to remote and hybrid working arrangements over the past several years means that almost all organizations will have business data stored on employee laptops and other devices that is not stored elsewhere.
Backing up data using multiple storage methods is different from backing up servers or endpoints using multiple storage methods. The first focus on the data, the second on backing up a particular manifestation of that data. If authoritative business data is distributed across multiple locations, the resilient data backup strategy in its defensive mode places primacy on backing up the data. In its recovery mode, on the other hand, backup processes must either enable the recreation of full operational systems with the latest data or work in conjunction with orchestration platforms to deploy data wherever it is needed.
REPORT CARD FOR RESILIENT DATA BACKUP (ADDITIVE)
- Baseline: All data is backed up using multiple storage methods, enabling restoration of any system or application. Restoration procedures are regularly tested and confirmed
- Better: Ransomware protections are integrated in backup and restoration procedures
- Best: Changing backup settings or deleting backup data is subject to elevated protections
6. Prepare to Recover After an Incident
While it is almost certain that every organization will face opportunistic attackers— and perhaps determined ones, too—it is not certain that every organization will become a victim of a ransomware incident. But if that does happen, having a response protocol or recovery plan approved by the Board and ready to enact is invaluable.
STEPS FOR RECOVERING AFTER AN INCIDENT
Essential steps for preparing to recover after compromise are:
Develop a recovery plan
If your organization suffered a ransomware incident, how would you recover? The extent and impact of a possible incident could range across multiple dimensions, hence plan for several compromise scenarios involving both users and devices, local and remote. Developing a recovery plan before an incident offers two benefits: first, developing a plan doesn’t have to be undertaken during the heat of a real incident, and second, identifying weaknesses in current protections enables earlier mitigation.
Check and confirm the “small” but critical details
Assess, test, and verify that the specifics are appropriate to the risk. If backups are essential for data restoration without paying the ransom, confirm the right backup media is being used and that advanced controls over backup settings have been activated (not merely licensed). If biometric identity controls are being introduced, ensure they are stored securely and subject to the right protections. These and other small details form the foundation for effective and prompt recovery following an incident.
Verify insurance coverage is fit-for-purpose
Cybersecurity coverage was a goldmine for insurance companies until ransomware payouts severely undermined profitability. Insurance companies offer different types of policies, and untested reliance on the wrong type has costly implications after a ransomware incident. An Australian company recently lost a court case against its insurance provider after claiming for financial damages from a ransomware incident under a traditional crime policy. The company had not taken out insurance coverage specifically for cybersecurity incidents.
Reduce the negligence scope
Locate and review previous cybersecurity readiness assessments undertaken specifically for your organization. Ensure identified weaknesses have been addressed or mitigated—otherwise a ransomware compromise through any of these will reek of internal negligence. If more than 12 months has elapsed since the last readiness review, commission an update.
REPORT CARD FOR RECOVERING AFTER AN INCIDENT (ADDITIVE)
- Baseline: Recovery plan developed, tested, and reviewed regularly
- Better: Test and confirm the details on which recovery is built
- Best: Conduct regular ransomware readiness assessments and act promptly on the recommendations, including for previous assessments
7. Prevent Abuse of Exfiltrated Data
Ransomware attacks are increasingly focused on exfiltrating data rather than encrypting data only, because exfiltrated data adds weight to the extortion used by threat actors against victim organizations when demanding a ransom payment. For example, threat actors tie the ransom payment to preventing the publication, sale, auctioning, or other usage of the stolen data—not for providing the decryption key. While organizations with resilient data backups can overcome operational disruption and avoid paying the ransom to gain a decryption key, and organizations with anti-data exfiltration and DNS monitoring solutions can stop data exfiltration in process, recovering from the theft of readable data is much more difficult. The only viable way to protect organizational data is through pre-exfiltration protection.
SOLUTIONS FOR PREVENTING EXPLOITATION OF EXFILTRATED DATA
Several core data disciplines need greater attention by organizations to prevent exfiltrated data from being exploited. For example:
Extend data discovery and data classification beyond sanctioned tools and cloud apps
Documents, email messages, and chat threads in unstructured communication and collaboration tools and cloud apps (e.g., Box, Dropbox, OneDrive, SharePoint, Teams, and Google Workspace) are routinely used by employees to exchange information about patients, customers, financial matters, and project details. In addition, cloud apps are being adopted by business units, teams, and individuals to circumvent legacy and restrictive sanctioned apps, creating data repositories that are not as well-protected as those under the purview of IT and security teams. Data discovery and data classification tools enable the identification of confidential, sensitive, and personal data that needs to be protected—irrespective of where that data is stored. Organizations that don’t have the visibility and optics into where their data resides cannot protect it.
Encrypt data—in-transit, at-rest, and in-use
The use of strong data encryption (i.e., stronger ciphers)—intentionally and under the control of the organization—assures data is safeguarded throughout its lifecycle, even if stolen by a threat actor. Encryption solutions are frequently, but not ubiquitously, used with data in-transit and data at-rest. Encryption is less frequently used with data in-use, for example, homomorphic encryption for secure computation. Most organizations have room for improvement in the use of encryption to secure their own data and prevent its exploitation if it is exfiltrated by a threat actor. Encryption and pseudonymization are two data protection technologies mentioned explicitly in the GDPR (General Data Protection Regulation) in the European Union as means of protecting personal data, and reducing the likelihood that a data breach can successfully compromise plaintext data values.
REPORT CARD FOR PREVENTING ABUSE OF EXFILTRATED DATA (MATURITY)
- Low: Sensitive, confidential, personal, and other regulated data is protected at rest and in transit
- Medium: All data is protected at rest and in transit
- High: All data is encrypted all the time, even during use
8. Develop Resilient Operating Processes
Ransomware attacks disrupt normal operating processes. Employees cannot rely on IT systems to accomplish their tasks in the normal way. In some industries, employees may be able to cease operations for several hours without causing significant loss. In other industries, every minute counts—such as healthcare, transportation/logistics, and law enforcement. Leaving it up to staff to scramble and improvise after a ransomware incident is a recipe for costly implications downstream.
SOLUTIONS FOR RESILIENT OPERATING PROCESSES
Developing resilient operating processes includes:
Having alternative operating processes on standby
Since IT systems will be inoperable after an incident, access to forms and manuals through alternative means will be required. For organizations where it matters, create commonly used forms and manuals for paper-based processes when IT processes are rendered inoperable. Offer staff training on a suitable cadence covering how to access and use these resources. Options include use of the organization’s multi-functional printing devices, agreements with local print shops, or USB thumb drives stored securely for use with new printers purchased under urgency after an incident.
Preparing to capture data created on paper-based forms
For organizations where it matters, data entered on paper-based forms during a ransomware incident, such as patient charts in a hospital, must be accounted for once normal IT systems are restored. Options for capturing data include scanning forms to be associated with a case, project, or patient, automatic conversion techniques to transform handwritten text into machine-readable text, human transcription of captured records, or submission of a photo of a form via a mobile app. Having traceability of actions, decisions, and authorizations during the use of alternative processes is just as important as when normal IT systems are in use. In some industries, not having all records available would be a violation of compliance requirements. In others, it could also be the basis for a disgruntled client or family member to sue for malpractice or negligence.
Maintaining back up methods for contacting staff and coordinating with clients
Scheduling systems are used for staff rosters, patient bookings, client meetings, raw material movements, and much more. If these systems are compromised through a ransomware incident, lists of upcoming events, and contact details for people will be inaccessible. Even the ability to call people to reschedule meetings or appointments—or recommend other suppliers—will be hampered for as long as systems are out of action, magnifying the chaos across the victim’s ecosystem. Options include priority restoration of scheduling systems, regular exports of upcoming events and contact details to a secondary storage system, or integration with air gapped devices receiving regular updates.
REPORT CARD FOR RESILIENT OPERATING PROCESSES (ADDITIVE)
- Baseline: Alternative operating processes developed, including required forms and manuals. Employees trained on how to work through significant system disruption
- Better: Close the loop for data developed/captured using manual processes
- Best: Minimize the scope for wider chaos by ensuring staff and clients can be contacted
We have examined eight areas for organizations to strengthen in anticipation of further ransomware attacks and incidents. Although attack indicators in 2022 have been lower than expected—due to sanctions against Russia following the invasion of Ukraine—ransomware gangs are extending their crime models, toolkits, and playbooks to inflict significant future damage. Strengthening protections and addressing current weaknesses across these eight areas puts organizations in a much stronger position when ransomware attacks increase again. In addition to reducing the likelihood of compromise at a given organization specifically, elevated protections also decrease the attractiveness and profitability of the ransomware crime model in general.