AI Assessment Disclaimer

Every result generated by MyCELPIP — including estimated CELPIP levels, Writing and Speaking dimension scores, pronunciation and fluency analytics, and Listening/Reading accuracy estimates — is an unofficial, AI-generated practice assessment. It is intended to help you understand your current performance and target areas for improvement. It is not an official CELPIP score, is not issued by CELPIP or Paragon Testing Enterprises, and is not a guarantee or prediction of your actual test result.

Listening and Reading results come from deterministic scoring against our own internally versioned conversion tables, clearly labeled as estimates. Writing and Speaking results combine deterministic metrics (word count, timing, pace, pauses), pronunciation-analysis output from a dedicated speech-analytics provider, and structured AI rubric evaluation — never a single, opaque AI call. All AI evaluations are validated against a defined schema before being shown to you, and every evaluation reports a confidence range rather than a single definitive number.

Results are always described using language such as:

• estimated practice level
• AI-generated practice assessment
• unofficial score estimate
• performance estimate
• preparation feedback

We do not describe results as:

• official CELPIP score
• certified CELPIP score
• guaranteed score
• same scoring system as official examiners

AI evaluation and speech-recognition technology can make mistakes — including misjudging borderline responses, missing context a human rater would catch, or mis-transcribing unclear audio. Results may vary between similar responses. Use your estimated level as one input among several (practice volume, mock-test performance, instructor or tutor feedback) when judging your readiness for the official test.

We periodically compare AI-generated evaluations against human-reviewed benchmark responses and adjust our evaluation prompts and rubric versions accordingly. This process improves consistency over time but does not make our system equivalent to official CELPIP examiners or their scoring methodology.