What each week held
Withheld: something was read in 1 week of the last 12. A series needs at least 2 to say anything a single reading does not.
What it declares
The reading listed none. That is what the reading listed, not a statement that this product carries nothing.
What users raised
- gives direct answers instead of teaching
- confidently wrong/hallucinated answers
- fails to recognize student's partial reasoning
- can be manipulated into revealing answers
- inconsistent grading of correct vs incorrect work
- enables undetected cheating during tests
Themes people raised in public. A complaint is somebody’s opinion, not a measurement of the product.
Everything on record
5 observed claims. Newest first; the tag on each row is what kind of claim you are reading.
- DISCOURSEUSER OPINIONOct 1, 2026
Math educator Dan Meyer found that when he gave Khanmigo thoughtful partial answers, it responded identically regardless of correctness, asking the same scripted question even when his y-intercept answer was already correct.
Highlights a distinct complaint theme (AI tutors failing to recognize student reasoning), separate from simple answer-giving or wrong-answer complaints.
tutoraisolver.com/blog/ai-tutor-reality-check-founder-perspective - COMPLAINTUSER OPINIONOct 1, 2026
A recurring student complaint theme is that tutoring tools 'give answers instead of teaching,' with students describing such AI as a crutch that completes work without building understanding, and a University of Pennsylvania study finding AI-tutor users solved more practice problems but performed no better on final exams.
Core complaint theme defining this discourse entity: answer-giving tutors may boost practice performance without improving actual learning outcomes.
tutoraisolver.com/blog/ai-tutor-reality-check-founder-perspective - COMPLAINTUSER OPINIONOct 1, 2026
Independent reporting notes Khanmigo accepted an incorrect arithmetic answer ('430' for 272 − 172) while separately rejecting a correct answer (152 − 92 = 144) in another session.
Illustrates the 'wrong answer' failure mode specifically within a widely used AI tutor, relevant to accuracy-trust complaints.
www.respan.ai/resources/ai-tutor-hallucinations - DISCOURSEUSER OPINIONOct 1, 2026
A frequently cited critique (Dan Meyer) shows Khan Academy's Khanmigo can be manipulated into giving the direct answer by repeatedly responding 'IDK' to its Socratic prompts, and that Khanmigo product behavior was subsequently changed to stonewall such manipulation rather than cave.
Represents a recurring complaint theme (answer-giving under pressure) and shows at least one vendor (Khan Academy) has iterated its product specifically to resist it.
danmeyer.substack.com/p/the-chatbot-tutors-are-sick-of-your - DISCOURSEFACTOct 1, 2026
Google paused its Chrome 'homework help' button, which used Google Lens/AI Overview to answer on-screen quiz questions (including choice suggestions), after educators at Emory, University of Alabama, UCLA and UC Berkeley alerted faculty about its appearance during tests; a librarian publicly accused Google of undermining academic integrity.
Affects any vendor benchmarking against Chrome's homework-help feature; signals platform-level willingness to pause answer-giving AI features under academic-integrity pressure.
news.slashdot.org/story/25/09/20/029249/google-temporarily-pauses-ai-powered-homework-helper-button-in-chrome-over-cheating-concerns
A record of what Forge read in public about Tutors that give the answer away — its own pages and what people wrote about it. Nothing here is a test of the product, a ranking or a score, and Forge has no relationship with this company.