负责任的 AI

以人为本的设计

We design Blue Yonder AI systems around human responsibilities and business outcomes. We aim to align automation, oversight, and safeguards with the level of risk to keep accountability clear.

AI Literacy and Responsible Usage

We promote AI literacy by making role-relevant information about our AI Systems available to users where appropriate. We encourage responsible use that considers expected benefits, potential risks, and best practices, while remaining aligned with the context and intended use cases for which our AI systems were designed.

Transparency and Explainability 

We aim to provide users with clear and appropriate information about Blue Yonder AI Systems, including when AI is being used. Where appropriate, we provide understandable information about relevant data sources, material factors, and human oversight behind significant AI outputs or actions so users can interpret and govern their use responsibly.

Fair, Unbiased, and Non-Discriminatory 

We design and operate AI Systems with the goal of supporting fair and non-discriminatory outcomes. We put in place proportionate measures, including tools, testing, and ongoing monitoring, to identify, assess, and reduce unfair bias in data, models, and processes across the AI lifecycle.

Data Security and Privacy 

We design and operate AI Systems with technical and organizational controls intended to protect customer data, consistent with customer agreements and applicable law. We aim to minimize data use to what is needed for the intended purpose and apply access, retention, and privacy protections proportionate to risk.

人工智能安全 

We build AI Systems with security and resilience in mind, including controls for AI specific threats where relevant. We use risk-based testing and vulnerability management along with detailed reporting and logging to assess and improve these controls over time.
 
Governance and Accountability
 
We use governance processes for AI Systems throughout the life cycle, including defined roles, review practices, and escalation paths. We evaluate risk to align the level of controls with the degree of impact.

Applicability

We offer AI Systems where they can deliver meaningful value for the use case and operating context. Before deploying or expanding automation, we assess fitness for purpose, limitations, and potential impacts.

Sustainability
 
We consider environmental sustainability in how we build and run AI Systems. Where AI-enabled optimization may improve efficiency, we aim to support reduced waste and resource use.

Trustworthy and Reliable
 
We design and build our AI Systems to perform reliably for their intended use through in house and real-world validation, quality controls, and continual improvement.