LEARN. PRACTICE. GROW.Concept and mechanism
The historical syllabus uses five Salesforce Trusted AI Principles: Responsible, Accountable, Transparent, Empowering, and Inclusive. These names belong to a specific framework; they should not silently be replaced by later lists for generative or agentic AI. Applying them requires observable decisions. Responsible involves appropriate use and protection of entrusted information. Accountable requires owners who respond to impacts and seek independent feedback. Transparent aims to make recommendations understandable, with factors and limitations suitable for the decision-maker. Empowering supports human ability to decide, correct, and work effectively. Inclusive considers affected people and contexts that creators may not represent. A principle is not a technical guarantee against error or harm.
Guided application
In a project, translate each principle into acceptance and follow-up criteria. Ask who can challenge a recommendation, who investigates a complaint, and how effects on different groups are measured. If data contains almost only day requests, involve the night shift before expanding the pilot. Removing a sensitive field does not automatically remove correlated information in other attributes. Review outcomes, proxies, and context with qualified people. A premortem helps imagine failures before service introduction by assigning actions, signals, and owners. The team should distinguish ethical analysis from specific legal conclusions; this course teaches management decisions and does not establish legal compliance for a particular implementation.
Example: a challenge form identifies the outcome, retains authorized context, and routes analysis to an owner with an agreed response time.
Common pitfalls
Memorizing names without controls; explaining without allowing challenge; assigning every impact to the vendor; treating a removed attribute as proof of fairness.
Related topics: Choose the AI task and quality evidence · CRM capabilities, recommendations, and actions · Human oversight and response to AI failures
A principle becomes useful when it changes a decision, control, or responsibility.
Reference: Meet Salesforce Trusted AI Principles · AI Associate historical Spring 24; retired 2026-02-02