How to use AI for business continuity and disaster recovery planning
Are you incorporating AI into your BC/DR planning? If not, there are multiple opportunities to include AI in the planning process, both before and after a disaster occurs.
Artificial intelligence has been a formal academic and technological discipline for over 60 years. Companies have developed many different approaches using the power of computers to simulate human intelligence capabilities, such as reasoning and inference. With the dramatic increase in computer processing power over the past 20 years, the ability to utilize all of this has made AI a much more realistic business opportunity.
When used for business continuity and disaster recovery planning, AI has the potential to change key processes for the better. Business continuity and disaster recovery (BC/DR), as well as incident response planning, could receive a boost in functionality and relevance from the introduction of AI.
First, let's examine where AI might be most beneficial to business and government organizations. The figure below shows a typical timeline for implementing AI for business continuity and disaster recovery planning:
This is a typical sequence of activities for planning, exercising, activating and continuing BC/DR and incident response plans. In each of the above situations, the role of AI is likely to be different. Let's take a closer look at each phase.
Plan development
IT teams might use an AI-enhanced plan for development, and its database coupled with some inference or cognitive functions might help planners make better decisions as to how to organize the plan. The AI component may be able to offer additional guidance, as in "do this, not that," which can make the plan more adaptable for a range of incidents.
AI for business continuity might be very useful for performing business impact analyses and risk assessments, especially when analyzing data from interviews and other documentation. It may also be possible to tweak the tool to examine the data for unique relationships -- e.g., interdependencies -- that may be hidden and, ultimately, significant.
Exercising plan
An AI component in an automated exercise system may be useful to introduce new and potentially significant situations into the exercise flow. As the exercise progresses, data entered into the AI-supported system can be used to trigger changes in the sequencing of activities.
This is the last opportunity to incorporate AI into the business continuity and disaster recovery planning sequence prior to a disruptive event taking place.
Incident response actions
In the immediate aftermath of an incident, it can take time to evaluate the event, determine its severity, and then make decisions regarding evacuation, system recovery and restart. If the incident response team has a smartphone with speech recognition and an AI component, team members could describe what's happening to the system and have the system offer recommendations for next steps.
While there may be no real substitute -- at this time -- to human experience and the ability to make decisions in an emergency, it may be valuable to have additional insights from AI-supported systems.
Clearly, such a system would need a considerable amount of experiential data in place from which to make its recommendations. While there may be no real substitute -- at this time -- to human experience and the ability to make decisions in an emergency, it may be valuable to have additional insights from AI-supported systems.
It may be possible to have the AI-supported system commence technology recovery activities if the system exceeds certain performance thresholds. This could be done in parallel with incident response activities, and it might save valuable time in terms of responding to a variety of situations.
Technology recovery actions
Assuming a technology disruption has occurred, an AI-based system could be programmed to initiate recovery activities, as noted above; connect with cloud-based backup and recovery resources to initiate system failover activities; and ensure that backed-up data, virtual machines, databases and other technology resources are in place and ready for activation.
The speed of launching these activities might be important to reduce the time it takes to recover and limit potential technology damage.
Business recovery actions
Depending on the severity of the incident, an organization may need to initiate various business-related actions. For example, it may be necessary to arrange an alternate workspace, activate alternate resources for manufacturing, and locate and deploy devices -- e.g., laptops, servers, scanners and printers -- at an alternate location and have them activated.
If an AI-supported BC system learns about many different situations, has a list of alternate resources, as well as a list of possible response and recovery strategies, it may be able to provide rapid recommendations for recovery.
Linking business and technology recovery
Another AI for business continuity and disaster recovery use, as shown in the graphic, is a connection between business recovery and technology recovery actions. If a technology incident occurs, it may not adversely affect the organization's ability to perform its normal work.
If a technology incident is more serious, such as a disruption to the internet or a massive power blackout, it may be necessary to trigger BC activities based on the technology outage quickly and efficiently. Assuming the BC/DR elements share AI resources, it may be possible to quickly launch the most appropriate sequence of activities.
Alternatively, the system could provide recommendations to its human counterparts for guidance, with the humans managing the actual response. Such a system might also be programmed to respond to nontechnology disruptions, such as a flu outbreak, an active shooter or a hazardous materials event.
The idea of a system activating other systems, launching response activities and communicating with emergency resources may be a bridge too far. Instead, it may be more helpful -- initially -- for an AI-supported system to provide guidance and recommendations to emergency teams.
Next steps
Once a company has managed an event and it is no longer an immediate threat to the organization, some systems might be capable of providing guidance and recommendations for recovering to a business-as-usual state.
Clearly, BC plans should be able to provide such direction, but given the wide variety of possible events, it may be useful to have access to more detailed data on how other organizations have handled similar situations, as well as statistical data on such events.
Organizations that are actively using AI-based systems may be able to enhance their business continuity and disaster recovery planning actions on their own. The key is to ensure that the business and its mission are protected.
