The worst thing that can happen to your AI endeavor, is that you end up damaging people’s lives through good intentions.

Without proper AI ethics practices, your clever AI application could be filtering the best candidates from job applications due to their gender, you could be preventing good citizens from qualifying for loans, or your AI application might even result in children unfairly being taken from their parents.

Extreme as some of these examples sound, they have all happened, impacting people’s lives in very bad ways. Yet they could have been prevented if AI Ethics best practices had been followed by the creators of these applications.

AI Ethics has many dimensions, including:

• Executive sponsorship and oversight
• AI Ethics committees that create and implement best practices
• AI and business professionals trained in AI Ethics
• AI platforms and tools that can automatically detect and fix bias, and explain the reasons behind AI-based recommendations to humans
• AI methods that enforce bias detection and explainability — this is where MLOps comes in

#data-science #ai #ai-ethics #ai-trust