Trustworthy AI: Taking a Trustworthy Approach to the AI and Analytics Life Cycle With SAS Viya
Organizations that embed AI solutions within their business processes also take on the responsibility of mitigating the inherent risks of AI.
AI models are trained on historical data and can perpetuate biases that were hiding in the training data. AI technology often uses sensitive data such as private personal information, health care data or business use cases for creating, training and using models. Unauthorized access to or inappropriate disclosure of these types of data can cause harm to individuals or organizations. Opaque AI systems can foster distrust by masking the decision flows and can increase the challenge of understanding the basis of a decision. Plus, they limit the auditability of a system.
All these issues can lead to AI systems that make unfair or discriminatory decisions, becoming a legal, reputational and ethical liability for the business. The unintended consequences resulting from not addressing and mitigating these risks pose real threats to organizations.