Cisco’s Annual Cyber Security Report 2018 provides an insightful account into the changing cyber security landscape. This article summarises some findings of the report pertaining to cloud security.
According to the report, increased security was the principle reason security professionals gave for organisations deciding to host corporate applications in the cloud.
Organisations who have a security operations team are likely to have a well defined cloud security approach that may include the adoption of Cloud Access Security Broker (CASB) as they deploy to the cloud.
Many smaller organisations however are adopting cloud services without a clear security strategy, there is therefore a blurring of the security boundaries where many organisations are not certain about where their responsibilities end and where the responsibility of the cloud provider starts.
Cyber attackers are increasingly taking advantage of this blurring of the boundaries to exploit systems.
An increasing trend amongst cyber attackers is to use legitimate cloud services to host malware and command and control infrastructure. Public clouds that have been used for malware activity include Amazon, Google, DropBox and Microsoft.
This makes it doubly difficult for security teams to identify bad domains and take protective measures without risking significant commercial impact caused by denying user access to legitimate business services.
The misuse of legitimate services is attractive to cyber attackers for a number of reasons;
Cyber attackers are effectively using legitimate and well known cloud infrastructure with their attendant benefits; ease of scale, trusted brand and secure features such as SSL. This enables them to scale their activity with less likelihood of detection if current protection methods are retained.
The challenges posed for the security teams defending organisations from these new threats call for a more sophisticated approach because in effect you need to block services that users are trying to access for legitimate work such as Amazon or Dropbox. Furthermore, the legitimate services are encrypted and so malware will be encrypted and evade most forms of threat inspection techniques– the threat will only become apparent after it has been activated on a host.
Intelligent cloud security tools will need to be deployed to help identify malware domains and sub-domains using legitimate cloud services. Such tools can also be used to further analyse related malware characteristics such as associated IP addresses, related domains and the registrant’s details.
An emerging and valuable approach to detect anomalous behaviour is machine learning.Machine learning algorithms can be used to characterise normal user activity, unusual activity can be identified, and action taken automatically.
To meet the range of challenges presented by cloud adoption,organisations need to apply a combination of best practices, advanced security technologies, and some experimental methodologies especially where they need to overcome the use of legitimate services by cyber attackers.