AI value chain actors
Using Saidot, you will be able to
Implement governance of both provider and deployer cases.
Identify and document your AI suppliers.
Invite your Partner to conduct AI governance utilising Saidot’s methodology and platform, including policy templates.
Publish system transparency report to one or more organisations.
Link your partner’s published AI system and other potentially relevant third-party components.
Apart from the roles and responsibilities allocated internally (Roles and Responsibilities in Governing AI), the organisation shall ensure that responsibilities within their AI system lifecycle are also allocated between the organisation, its partners, suppliers, users/deployers and third parties, in order to ensure that the organisation comprehends its responsibilities and remains accountable, and risks are appropriately apportioned when third parties are involved at any stage of the AI system lifecycle.
Examples of well-defined roles of AI value chain actors along the AI system lifecycle can be divided into primary roles and supplier roles as follows:
Type | Role | Description |
|---|---|---|
Primary roles | AI provider | An organisation or entity that provides products or services that use one or more AI systems. |
AI deployer | An organisation or entity that uses AI products or services. |
Type | Role | Description |
|---|---|---|
Supplier roles | AI model provider | An organisation or entity providing models to be integrated into AI products or services. |
AI system integrator | An organisation or entity that integrates AI components into larger systems, potentially also including non-AI components. | |
Data provider | An organisation or entity that provides data used by AI products or services. | |
AI developer | An organisation or entity that develops AI services or products. | |
AI auditor | An organisation or entity that is auditing organizations producing, providing or using AI systems, to assess conformance to standards, policies or legal requirements. | |
AI evaluator | An organisation or entity that evaluates the performance of AI systems. |
*Sources: Adapted based on ISO/IEC 22989:2022(E), ISO/IEC 42001:2023(E), EU AI Act