PEGACPDC24V1 practice question 32 of 35

Which two elements ensure fairness in AI-driven action prioritization? (Choose two.)

Full set: 350 questions
Answer options
A: Transparent evaluation criteria Correct answer
B: Limiting recommendations to specific channels
C: Dynamic updates based on customer data Correct answer
D: Standardized action templates
Correct Answer:
A. Transparent evaluation criteria; C. Dynamic updates based on customer data
Explanation

Fairness in AI-driven systems, such as Pega Customer Decision Hub, relies on transparency and adaptability. Transparent evaluation criteria (A) are essential for auditing the AI's decision-making process. This allows organizations to understand why certain actions are prioritized, identify potential biases against specific customer segments, and ensure the logic aligns with ethical guidelines. Dynamic updates based on customer data (C) allow the adaptive models to continuously learn from real-time interactions. This ensures that prioritization is not based on static, potentially biased historical data but on current, relevant customer behavior, thereby promoting equitable and fair treatment over time.

Why Incorrect

B. Limiting recommendations to specific channels: This is a business or delivery strategy, not a fairness mechanism. It can introduce bias by excluding customers who do not use the selected channels.
D. Standardized action templates: This ensures consistency in the presentation of an offer but has no bearing on the underlying AI logic that determines who receives the offer and why.

References

1. Pega Academy: In the Customer Decision Hub Foundation (24) course, the "Responsible AI" module emphasizes transparency as a key pillar of ethical AI. The "Ethical AI" lesson explains how Pega's capabilities provide transparency into AI-powered decisions to detect and mitigate bias.

Source: Pega Academy, Course: Customer Decision Hub Foundation (24), Module: Responsible AI, Lesson: Ethical AI.

2. Pega Documentation: The documentation on "Ethical bias policy" explicitly discusses the need for transparency and monitoring to ensure fair outcomes. It states, "To ensure that your business strategy is ethical, you can create and use an ethical bias policy that detects bias in your models and mitigates its impact." This directly supports the need for transparent criteria.

Source: Pega Documentation, "Creating an Ethical bias policy," Section: About ethical bias policies.

3. Pega Documentation: The documentation on "Adaptive models" describes how they "continuously and automatically learn from customer interactions." This dynamic updating (C) is crucial for fairness, as it allows the system to adjust its predictions based on the latest data, rather than relying on potentially outdated and biased static models.

Source: Pega Documentation, "Adaptive models," Section: Adaptive model rules.

Scroll to Top