Using AI in Risk Management—Opportunities and Risks

AI presents a risk management dual-edged sword: it’s both a powerful tool for managing risks and something that itself requires risk management. Understanding both sides is crucial for project and program managers. Here are key considerations when using AI to enhance your risk management.

The Opportunities

• AI can enhance the approach to risk identification and monitoring. Loading your business case and risk registers into an LLM can help identify risks you may not have considered. But it goes deeper than that. For executives and portfolio managers overseeing multiple projects, AI can analyze patterns in risk registers to identify where similar risks appear across projects. A risk categorized as low-impact in isolation might become critical when it surfaces across three or four projects.
• AI helps with risk register management. AI can be used to flag risks that haven’t been recently assessed, action items that have expired, or triggers that indicate certain risks require immediate attention. It can create mind maps that show interdependencies across projects and help identify when project-level risks should be escalated to program or enterprise levels.
• Multiple AI platforms that can be leveraged. Use multiple AI tools with the same prompt when analyzing risks, then compare the results. Different tools will produce varying results. Prompting why one tool identified something others missed can deliver valuable insights. Ask one AI to generate prompts for another AI to optimize the risk-related information you receive.

The Risks

• Sharing confidential corporate information in public AI environments can be problematic.Ensure appropriate safeguards are in place. Use privacy settings or set up your own internal AI environment.
• AI tools can take things out of context. Always verify AI outputs. Tools like Perplexity provide footnoted sources, making this easier. They allow you to check whether the AI has taken information out of context or misinterpreted source material.
• Data integrity must be managed. Your AI is only as good as the data it’s trained on. If you’re relying on AI to help make decisions but the underlying data has integrity issues, you won’t make much progress. Ensure you understand how the AI’s data was built, and how it will be kept up to date. Watch for data with multiple intervention points, unclear data stewardship, or unreliable transformations. With these issues, your AI will produce unreliable and possibly detrimental results.

Interestingly, the biggest risk might be not using AI at all. This technology is the direction business is heading, and familiarity with AI tools is becoming essential for sponsors, project managers, and program managers. Failing to develop AI competency is a risk that will become an issue as the technology continues to advance and reshape how organizations operate.