Responsible AI Principles
Microboard intends to build responsibility into the architecture of AI Computing.
Principles
Human-centered purpose
AI should support people and organizations with human oversight appropriate to risk and context.
Privacy
Personal data used in AI capabilities must be handled in accordance with the LGPD and applicable data-protection requirements.
Security
AI capabilities should be assessed for system, data, firmware, software, access and foreseeable misuse risks.
Transparency
Where relevant, users should receive understandable information about AI capabilities, limitations and use.
Robustness and operational safety
Capabilities should be tested proportionately to risk and be capable of update, correction, limitation or safe deactivation where needed.
Non-discrimination
Microboard will seek to reduce risks of unjustified bias and discriminatory impact in uses under its control.
Intellectual property and data
Integrations should consider rights in data, software, models, content and intellectual property.
Accountability
Roles of suppliers, Microboard, customers and users should be defined according to the solution.
AI supply chain
Third-party models, components, software and AI services should be assessed according to risk, security, privacy, licensing and intended purpose.
Communicating capabilities
Microboard communicates AI capabilities based on the actual hardware and software configurations of its products and platforms.