Ten Simple Questions for an Effective Discussion of Information Security
Information security should be one of the most important risk areas of focus for boards. However, according to the 2017–2018 NACD Public Company Governance Survey, 88 percent of surveyed directors indicated that they had only some or little knowledge about how to navigate cyber risk. It’s clear that too few directors feel qualified to have this conversation in any degree of depth.
When I joined Amazon.com in 1998, Jeff Bezos, the company’s CEO and chair, viewed security as the most threatening, potentially company-ending risk that the company faced. Since then, many companies have elevated security risk to their technology, the infrastructure on which they depend, as the greatest existential threat to their enterprise. Yet boards struggle to quantify these risks, to determine their tolerance for security risks, and to assess the company’s security program.
In their discussions of security risk, security leaders and board members are constrained by time, frame of reference, shared vocabulary, experience, and understanding of the adversary. Board members could use some help.
I propose ten simple questions that could enable discussion, provide board members with a lens through which they can broadly view the company’s security program and posture, and prompt security leaders to build a shared understanding of the company’s risk profile, threat landscape, and most important security initiatives.
1. Who is in charge?
It is critical for the board to identify the most senior information security leader in the company. This should be a person explicitly designated to lead the program, with the requisite skills, resources, and authority to execute it. This person commonly goes by a title such as chief information security officer (CISO), chief security officer, or head of security, among other titles. Sometimes, companies will take a tiered approach to security. In such cases, the leader of the security team plays a pivotal role, and the board needs to be comfortable that their position and authority is consistent with the importance that the board places on security.
If you identify someone who has security as one responsibility among a portfolio of others, it’s necessary to determine who has single-threaded focus on information security. Once that person is identified, you can discuss whether they have the proper ownership and resources to go with the responsibility, their reporting chain, the support that they receive from the rest of the company, and their relationship with the board. Regardless of who they directly report to, this person should be accountable to the board.
2. How do we assess risk?
Security is about risk management. It’s critical for directors to understand the process of identifying and analyzing security risks, how their likelihood and impact are estimated, how the appropriate controls are prioritized and implemented, how their efficacy is tested, and how results are monitored. Some potential security events are low probability and extremely high impact, making it more difficult to compare them to other risks. Nevertheless, it’s critical to go through the exercise of determining risk appetite, assessing and qualifying risk, quantifying overall exposure, and placing it within the company’s overall risk management framework. Finally, it’s important to be candid about your confidence in the risk assessment.
3. Are we focused on attacks?
It’s important to focus on managing the most critical threats and on breaking the attack kill chain—the structure of an intrusion—rather than to engage in “security theater,” or activities that give the appearance of competence while lacking in substance. Budgets are limited and security talent is in very short supply, so resources should be focused on establishing an architecture that has sufficient defense in depth, resilience, and intelligence to survive modern attack types.
Traditional approaches to defensive security that were dependent on protecting the perimeter of the enterprise continue to prove insufficient. Today, defenders must understand the adversary’s attack mechanisms, work backwards from the path of the attack, layer defensive measures throughout the enterprise, intervene before the attacker can extract sensitive data, and teach employees and customers to play their crucial part.
4. What’s our most important asset?
This question shouldn’t take long to answer. It should drive a discussion between the board and the security leader about how data and services are classified, the policies that are established for their defense, and the required and recommended controls for each class. When a new service is established, this classification framework in combination with the new service’s threat model should make it relatively easy to decide who is responsible for mitigating threats and what controls should be put in place.
When asked to rank their biggest cybersecurity fears, 41 percent of directors said they are most worried about brand damage. While customer trust is the key asset in many businesses, it’s important to identify the specifics of what would be the most devastating loss for the company. It’s only then that a thorough, qualitative assessment of the most critical components of the security program can occur.
5. How do we protect our most important asset?
Board members can calibrate the overall risk profile of a security program once they understand how the most precious asset is protected. The answer to this question should discuss the high-level threat model for that most important asset and, in the context of modern attack patterns, the mechanisms used to defend it. The answer should reflect that this is a journey on terrain that is shifting. There should be an iterative process of quantifying the risks of different threats, and of mitigating the most significant ones.
6. What’s our biggest threat?
This question forms the heartbeat of the conversation between the board and the security leader. It provides an opportunity to describe the company’s current security posture and its target state, and to refresh the board on the evolving threat landscape, the lessons to learn from emerging attacks, and the measures that the company is taking to mitigate the threats. For many companies, security risk is sufficiently important to warrant a discuss of this question at every board meeting, perhaps with a summary of the threat models for any major new products or services, and a review of the most significant risks at any recently acquired companies. When board members hear grandiose plans to address the biggest threat, but the deliverables are more than 18 months away, they may wish to ask for approaches to improve today’s posture without necessarily derailing the long-term solution. Don’t make the perfect the enemy of the good.
7. What do we control?
The board should assess the degree to which the company’s security policy and practices are explicit and prescriptive. Board members should be very suspicious of a security leader who claims to have complete control of the technology platform and the tools that employees use. Full control is usually a dangerous illusion, and any autocratic attempt to achieve it can lead to inflexibility and to employees working against or around the security program. Security should be viewed as a collective responsibility, rather than as a fixed constraint. Boards spend time assessing internal controls that for example provide confidence in custody over sensitive data and in the accuracy of financial reporting. Effective security leaders will distinguish between controls and control, and will strive towards “getting to ‘yes,’” rather than being the one who always says no. Getting to yes is easier if employees buy into a decision and if the path of least resistance is for them to do the right thing by default.
8. Are incident response and recovery plans tested?
This is one of those questions to which the answer can be “no” at most once. In the common case this question will lead to a review of responses and recovery from real incidents, in addition to a summary of simulated attack exercises, consideration of the fidelity of such exercises, and lessons learned. It provides the board with a view of the company’s capabilities in communication, response planning, incident analysis, risk mitigation under duress, and leadership.
9. Would we know if we’d been compromised?
Security technology vendors may tout breakthroughs that provide the ability to identify and prevent attempted compromises with perfect precision and recall. An effective conversation between a security leader and a board will take as a given that all attacks can’t be identified and prevented, and that compromises may already lurk undetected. This should lead to a discussion of actions to make prevention as strong as possible, to improve the probability of detecting lurking intruders, and to reduce the likelihood that they reach critical assets and extract them.
In a world where the edge of the company’s technology footprint is increasingly blurred, where the sophistication of attacks outpaces security awareness, and where advanced persistent threats are used by adversaries, it’s inevitable that the answer to this question will be nuanced.
10. Who would be told, and how do we expect them to respond?
Communication is a key part of a successful incident response plan. Each person, including the board, needs to know his or her role in communicating about incidents internally and externally. The question goes beyond incident handling to include recovery processes and the proactive management of any reputation impact that may arise from the incident.
As a board member, it’s worth thinking about two questions that I used back in 1998 to get Bezos thinking about his role in incident response:
- In the event of a high-severity security incident, do you think you’d be told?
- Would you like to be told?
Response and recovery go hand in hand. It’s tempting to avoid putting significant effort into planning for recovery from a major security incident, and while everyone would prefer to focus on prevention efforts with a goal of zero incidents, the reality is that there’s no such thing as perfect security. The recovery plan is part of responding to the incident, learning from it, managing communications, and getting the company back in business. A well-executed recovery plan has the potential to limit the reputation damage caused by the event, and to help management and other stakeholders to move beyond it.
Finally, a bonus credit question: Do you have the team and the budget that you need to be successful in managing the company’s security risk?
These 10 questions are a starting point for a longer conversation. Directors and the security leader should regularly employee a more thorough framework, such as the NIST Framework for Improving Critical Infrastructure Cybersecurity, to begin building a deeper understanding of their company’s security posture. While the NIST framework goes to considerably more depth, these 10 questions are intended to get to the essence of what is most important for a board to periodically review.
Tom Killalea (@tomk_) is a director of Capital One Financial Corp., MongoDB, Carbon Black, and Orreco. From 1998–2014 he served in various leadership roles at Amazon.com, including vice president of technology and CISO. All opinions expressed here are his own.