Introducing Collaborative Automated Course of Action Operations (CACAO): An Emerging Cybersecurity Standard to Quickly Define and Share Playbooks

June 17, 2020
| |
3 min read

A common challenge security teams face is ensuring their organization can continue to operate despite cyber attacks — in other words, to demonstrate their cyber resilience. Research from the Ponemon Institute acknowledges that the volume and severity of cyber attacks continues to rise, but suggests that organizations can improve their cyber resilience by a combination of process and technology improvements. The research said a key area of focus should be adoption of company-wide cybersecurity incident response playbooks to guide a business through its response to common attacks. In general, the need to accelerate and automate response playbooks has resulted in a growth in the adoption of security orchestration, automation and response (SOAR) tools.

A lack of interoperability and common playbook standards have slowed down the adoption of automated response, because incident response playbooks have had to be restricted to specific teams or technology implementations. However, this is an area that the technology industry has been working to address through the development of a new standard, led by the Organization for the Advancement of Structured Information (OASIS).

OASIS Collaborative Automated Course of Action Operations (CACAO)

The OASIS Collaborative Automated Course of Action Operations (CACAO) for cybersecurity standards can enable teams to quickly define and share playbooks. The idea of a playbook is to ensure that all members of both an organization’s security and business teams have clarity of responsibility and understand how to prevent, mitigate or respond to a security incident. Today there is no standardized way to document and share these playbooks across organizational boundaries and technology solutions. Since the launch of this OASIS initiative in September 2019, industry leaders from Accenture, Broadcom, Looking Glass Cyber Solutions, Cisco, Cyware, EclecticIQ, FireEye, Fornetix, IBM, New Context, Syncurity, ThreatQuotient, U.S. NIST and others have been collaborating to develop a standard to do just that. To date, the technical committee has created a working draft expected to move to a Committee Specification Draft (CSD) in the next few months.

The OASIS CACAO standard helps define an executable course-of-action playbook model for cybersecurity operations.  Specifically, the standard will offer organizations the ability to:

  • Author course-of-action playbooks and their processing instructions in a defined, machine-readable format (JavaScript Object Notation (JSON).
  • Reference or include other playbooks as part of a playbook they are authoring, or allow for composition of playbooks from smaller, more modular playbook components.
  • Digitally sign and provide opportunities to securely share or consume the playbooks. Those can be accomplished following the standard, both across organizational boundaries as well as technological solutions, while respecting “data markings” – a construct that represents the restrictions, permissions and other guidance for how the playbook can be used.

The Structure of a CACAO Playbook

Each type of CACAO cybersecurity playbook (investigation, prevention, mitigation, remediation) is proposed to be composed of five main areas along with an area to handle extensions. Those five areas are Metadata, Workflow, Actions, Targets and Data Markings, and the sixth is named Extensions.

Metadata – an area where information describing the playbook, or its properties, can be defined.  This information would include the playbook type, description, identifier, creator, version etc.

Workflow – this is the area where the logic for the playbook processing steps is defined. It may contain a collection of one or more actions providing guidance on how to address a particular security event, incident, attack or compromise. CACAO playbooks can refer to other CACAO playbooks and can be used like building blocks to compose comprehensive workflows.

Actions – collections of actions and commands to be executed are referenced as part of a processing step in a workflow.  Commands could be manual, an HTTP API, SSH, bash or openc2-json – the OASIS Open Command and Control standard.

Targets – collections of targets, such as the intended humans or devices required to execute commands, get defined in this area.

Data Markings – collections of data markings that define how the playbook may be used, shared or protected.

Extensions – this is the area where a playbook creator can define additions that get used in the current playbook.

Taken together, these areas document the playbook in a standardized way so that it may be easily shared across organizational boundaries and technology solutions.

Elevate Security Orchestration with CACAO

Despite the variety of known challenges facing security organizations today, the one thing that those organizations all have in common is a desire to improve their security programs. There is no such thing as “mission accomplished” for security programs, as organizations continue to face increasing volumes of security events and ever decreasing timeframes for threat mitigation and response. The attack surface continues to grow, and the shortage of security talent remains a challenge. As organizations evolve their security programs from manual to reactive to more automated and proactive approaches to threat management, the timing for a standard to accelerate the definition of executable playbooks and workflows could not be more perfect.

To find out more about this emerging standard join the OASIS webinar, “Elevating Security Orchestration with CACAO” on June 23, 2020, with presenters from LookingGlass Cyber Solutions, Broadcom, IBM, and New Context or visit the OASIS CACAO website.

Stephanie Hazlewood
Senior Technical Staff Member (STSM), IBM Security

Stephanie Hazlewood is a Senior Technical Staff Member (STSM) at IBM. As part of the IBM Security CTO team she works closely with offering management and cli...
read more