Systems and methods for screening and scoring outputs generated by artificial intelligence language models, including receiving a generated output in response to a user query, detecting errors by analyzing the generated output using a plurality of detection engines, each detection engine evaluating an error dimension, classifying each error into a severity level; assigning color-coded visual indicators to the errors, each color-coded visual indicator corresponding to a different severity level, the assigned color-coded visual indicator indicating the severity level of an error having the highest severity level, generating remediation guidance to remediate an error, and presenting screening results including the color-coded visual indicator, the errors with their respective severity levels, and the remediation guidance.