Addressing Demographic Bias in AI for Salesforce AI Associate

• A.Use a single centrally-trained model with fallback to human agents for non-English queries
• B.Use a translation layer to convert all inputs to English before processing
• C.Disable the bot in regions with dialect issues
• ✓ D.Deploy separate models fine-tuned on each dialect with centralized monitoring

Why D: Option D is correct because fine-tuning on local dialects improves accuracy and fairness, while centralized monitoring ensures oversight. Option A is wrong: fallback to human agents is good but still may cause delays and dissatisfaction; the model itself is not inclusive. Option B is wrong: translation may lose nuance and introduce errors. Option C is wrong: disabling the bot denies service and is not inclusive.

• A.Deploy the model but add a disclaimer that it may be less accurate for female customers.
• B.Deploy the model as is, because it still meets the overall accuracy threshold.
• ✓ C.Investigate the cause of the disparity, retrain the model with more representative data, and re-evaluate fairness.
• D.Manually adjust the model's output to ensure equal churn predictions across genders.

Why C: Option C is correct because investigating the cause of the disparity and retraining the model with more representative data addresses the root of the bias, which is the ethical approach. Option A is wrong because simply adding a disclaimer does not resolve the underlying unfairness. Option B is wrong because deploying the model despite known gender bias perpetuates discrimination. Option D is wrong because manually adjusting predictions introduces subjective alterations and may not be reliable or fair.

• ✓ A.The training data is imbalanced, which may cause the model to be less accurate for female leads, leading to unfair prioritization.
• B.The model will be biased against male leads because they are overrepresented.
• C.The dataset is too small to build a reliable model.
• D.Using historical data is unethical because it may not reflect current conditions.

Why A: Option D is the primary ethical concern. Using historical data without considering that it may reflect outdated or biased conditions can lead to unfair outcomes. While the imbalance in gender representation is a concern, it is secondary to the fact that the historical data may not accurately represent current or future conditions, thus potentially perpetuating past biases. Option A is a valid secondary concern but not primary; Option B is incorrect because overrepresentation does not inherently cause bias against the majority; Option C is irrelevant as dataset size is not identified as an issue.