What is AI Risk?

AI risk refers to the potential for an AI system to produce outcomes that are harmful, unfair, unreliable, or non-compliant. Every AI system - whether it is making decisions, generating content, or automating processes - carries some level of risk that needs to be understood and managed.

Why AI risk matters

AI systems are increasingly being used to make decisions that affect people, businesses, and regulatory outcomes. A model that shows bias in hiring decisions, a chatbot that leaks sensitive information, or an AI system that produces unreliable outputs can all create serious consequences for your organization.

Without a structured approach to understanding risk, these issues often go undetected until something goes wrong. Our platform helps you get ahead of that by building risk assessment into your AI governance lifecycle from the start.

How we define AI risk

We evaluate AI risk across six core dimensions. Each dimension looks at a different aspect of how an AI system could fail or cause harm. Together, they give you a complete picture of your AI system's risk profile. You can explore each dimension in detail in our article on The Six Dimensions of AI Risk.

• Bias - Whether the system treats different groups of people fairly and without discrimination
• Robustness - Whether the system performs reliably under varying conditions, unexpected inputs, or adversarial attacks
• Transparency - Whether the system's decisions can be explained, traced, and understood by stakeholders
• Privacy - Whether the system handles personal and sensitive data appropriately and in compliance with regulations
• Efficacy - Whether the system actually delivers on its intended purpose and produces useful, accurate outputs
• Exposure - How broadly the system is deployed, who interacts with it, and how likely identified risks are to be triggered in practice

How we assess risk in our platform

Our risk assessment process starts with Risk Mapping - a structured questionnaire built into our platform that evaluates your AI system across all six dimensions. Risk Mapping uses a layered approach with guided questions to build a complete risk profile for each AI system in your inventory.

From there, your team can run deeper, targeted assessments for each risk dimension. These assessments include both qualitative evaluations - where your team answers structured questions about the system - and quantitative testing, where we run automated checks against your system's actual behavior and outputs.

All of this feeds into a unified risk view on each Asset in your inventory, so you always know the current risk status of every AI system your organization manages.

Where AI risk fits in our platform

AI risk assessment is part of the Protect module of our governance lifecycle. Once you have identified and catalogued your AI systems through AI Discovery, the next step is understanding what risks they carry. Our risk and assessment tools give you the structured process to do exactly that.

Risk results also feed into workflows, compliance reporting, and mitigation planning - so risk assessment is not a one-time activity, it is an ongoing part of how you govern AI responsibly.