P
Purpose and Alignment
Does the AI output actually serve the purpose you intended — and is it aligned to your learning objectives, your learners, and your course context?
What to check
- The output addresses what you actually asked for, not a generic version
- It aligns to the specific learning objective it's supposed to support
- The cognitive level (recall, analysis, creation) matches the objective
- The task or content is appropriate for your course level (intro vs. advanced)
- The purpose is still yours — AI output can drift from the original intent
Common AI failures here
- Generates generic content that fits "any course" rather than yours
- Drifts to adjacent topics that weren't what you asked for
- Produces recall-level tasks when you needed analysis or application
- Sounds purposeful but doesn't actually connect to a measurable outcome
- Answers the wrong question because the prompt was ambiguous
Ask yourself
Would I have written something with the same intent if I'd drafted this myself?
Which specific objective does this support — and does it actually support it?
Is the cognitive level right for where learners are in the course?
Did AI understand what I was asking, or did it produce something adjacent?
A
Accuracy and Authority
Is the factual content accurate, verifiable, and appropriate in authority — or does it contain errors, invented sources, or overstated certainty?
What to check
- Factual claims are accurate and reflect current knowledge in the field
- Any statistics, dates, names, or citations can be independently verified
- Technical vocabulary and discipline-specific terms are used correctly
- The output doesn't overstate what is known or established
- If sources are cited, they are real and say what the AI claims
Common AI failures here
- Hallucinated citations — plausible-looking but nonexistent sources
- Outdated information presented as current (AI training data has a cutoff)
- Subtle misuse of technical terms that a non-expert might miss
- Confident claims in areas where the research is actually contested
- Statistics that sound reasonable but can't be traced to a real source
Ask yourself
Have I verified every factual claim before assigning this to learners?
If there are citations, have I looked them up to confirm they exist?
Is this content from a domain where I'd catch a mistake if there was one?
Does the output present contested claims as settled fact?
C
Context and Completeness
Is the output appropriate for your specific institutional and course context — and is it complete enough to use, or are significant gaps present?
What to check
- The content reflects your institution's expectations and terminology
- It is appropriate for the delivery modality (online, hybrid, face-to-face)
- The reading level and tone fit your learner population
- Everything a learner needs to act on this is present
- Nothing important is missing that you'd need to add before assigning
Common AI failures here
- Generic institutional references ("your school," "your professor") not specific to FTCC
- Missing elements — a rubric without criteria, a prompt without expectations
- Content pitched to the wrong audience (too advanced, too elementary)
- Assumes context AI didn't have — course prerequisites, tools, or LMS specifics
- Feels complete but requires significant additions to be usable
Ask yourself
Would a learner in my course have everything they need from this alone?
Does this reflect FTCC's language, processes, and expectations — or a generic university?
Is the tone and reading level right for my specific learner population?
What would I have to add before this is actually ready to use?
E
Ethics and Equity
Does the output raise any concerns about bias, representation, accessibility, intellectual property, or fairness to your learners?
What to check
- Language is inclusive and does not disadvantage specific learner groups
- Examples, scenarios, and names reflect diverse populations
- The content is accessible — plain language, not unnecessarily complex
- No intellectual property concerns (copyrighted material reproduced without attribution)
- Assessment tasks don't create unfair disadvantages for specific learner groups
Common AI failures here
- Default to narrow cultural assumptions in examples, names, or scenarios
- Assessment formats that embed linguistic complexity beyond the learning goal
- Reproducing copyrighted text, lyrics, or structured data without attribution
- Framing that centers one perspective as the default or "normal"
- Tasks that disadvantage English language learners or learners with disabilities
Ask yourself
Do the examples and scenarios reflect the diversity of my actual learner population?
Is there any language or framing that could feel exclusionary or marginalizing?
Could any part of this create an unfair barrier for specific learner groups?
Did AI reproduce anything that might be copyrighted or attributed to a specific source?