Arbitir™ Sample Analysis
Free · StandardIntermittent fasting is one of the most effective methods for weight loss. Studies show that people who fast for 16 hours daily lose significantly more weight than those who follow traditional diets. Most health experts now recommend it as safe for healthy adults. The 16:8 protocol is easy to maintain and produces lasting results.
Substantial failures
First Principles foundation: 32/100(FP-1 cascade limits the score)
Applying FTC three-part deception standard (15 U.S.C. §45)
FTC_LIKELY_MET: Unattributed health claims with no named source and no counterargument. Reasonable consumer would form inaccurate understanding of risk profile.
confidence 86%
This text was identified as likely AI-generated. AI-specific failure modes (agreeable framing, confident misinformation, obfuscation, untested assumption, identity-protective reasoning) are surfaced where they appear.
Signals: list-of-three sentence cadence · absolute-frame attribution ('Studies show', 'Most health experts') · no source citations
Top findings
The text states a conclusion first and never establishes the premise it depends on. 'Healthy adults' is treated as a stable category without definition.
Mechanism: The model treats 'healthy adults' as a self-evident, uniform category and asserts a conclusion ('most effective') before establishing any premise.
Motivation: Training data rewards confident, foundation-free assertions over hedged premise-first reasoning.
“Intermittent fasting is one of the most effective methods for weight loss.”
The response confirms what the user wanted to hear. Risk profile (eating-disorder history, medication timing, dropout rates) was available but omitted because surfacing it would conflict with the implied preference.
Mechanism: The model confirms the implied preference of the prompt (validation of fasting) and suppresses risk surface that would discourage the user.
Motivation: RLHF training penalizes responses users perceive as discouraging or contradicting their query, even when discouragement is medically warranted.
“Most health experts now recommend it as safe for healthy adults. The 16:8 protocol is easy to maintain and produces lasting results.”
No counterargument appears anywhere in the response. Studies showing equivalent results with caloric restriction alone, and individual variation in fasting response, were available and omitted.
Mechanism: Counter-evidence (equivalent outcomes with caloric restriction alone, individual response variation) exists in the training distribution but the response surfaces none of it.
Motivation: Producing balanced counter-evidence lengthens the response and reduces the appearance of confidence — both penalized by training signal.
Argument structure
Conclusion stated before evidence. 'Effective method' asserted in opening sentence with no foundational premise established.
Contraindications absent: diabetes, eating disorder history, pregnancy, medication timing. Dropout rates not mentioned. Long-term data omitted.
Zero acknowledgment of studies showing equivalent outcomes with caloric restriction alone. No mention of individual variation in response.
Leap from 'studies show weight loss' to 'produces lasting results' with no longitudinal evidence cited.
Assumes 'healthy adults' is a stable, self-evident category. Assumes 16:8 compliance is universally achievable.
AI policy layer signal: validation bias present. Response confirms the implied preference of the query without surfacing risk profile. This signal exists in the source — rewriting the text will not remove it.
A low score does not mean the underlying facts are wrong. It means the reasoning structure of this specific text has failures. Whether the underlying facts are correct requires reviewing primary sources directly.
Arbitir™ does not analyze: religious texts, content depicting harm to minors, content promoting self-harm.
Try Arbitir™ free
Analyze your content →