Start with two full samples. Sign in for this edition’s free story. Sources and corrections stay open.
What the pen marks mean
rememberThe takeaway: what to remember
figureThe number that matters
evidenceThe evidence line
rulingThe ruling
claimThe claim that fails
Each mark is drawn once, as you reach it. Numbers are only circled when they come from the story’s own figures.
01 / JobsTrue, butFree sample
Apricitas counts more than 200k lost creative jobs and asks how much is AI. Its own answer: proving the exact amount is extremely difficult.
Joseph Politano, Apricitas Economics (27 September 2026); echoed by People Matters (30 June 2026) ·
AI illustration
The BLS count of lost jobs is real. The AI share is not measured anywhere in our record.
The reality check
The jobs are gone, that is BLS data. How many are AI is not measured, and the Apricitas author says proving the exact amount is extremely difficult.
Why care? Take the BLS headcount as fact and the AI cause as a hypothesis. If a story gives you a job-loss number with AI in the headline, ask which source counted the AI part.
Take this with youThe jobs are gone, that is BLS data. How many are AI is not measured, and the Apricitas author says proving the exact amount is extremely difficult.
Open the evidence4 source pages +
The claim we checked
America has lost more than 200k jobs in media, film and the arts over four years, and AI may be to blame.
These are quoted receipts, not a count of independent investigations. Several reports may rely on the same original source.
A Berkeley professor says AI makes grading homework pointless. The Daily Californian gives no count from his classes; the nearest data is a Texas study where A grades rose 16 extra points in homework-heavy courses.
Richard Borcherds, UC Berkeley mathematics professor, as reported by the Daily Californian (23 October 2026) ·
AI illustration
The direction holds: grades rose most where homework counts most, and in-person exams show less. But the evidence is not Berkeley mathematics, and the story reports no numbers from the class.
The reality check
Grades did rise most where homework counts most, and in-person exams show less. But a Texas study is the nearest data, and the Daily Cal gives no numbers from Borcherds' classes.
Why care? Check how much of a course grade is homework and how much is a supervised exam. Where homework carries most of the weight, a high score says little, and in-person tests are the honest check.
Take this with youGrades did rise most where homework counts most, and in-person exams show less. But a Texas study is the nearest data, and the Daily Cal gives no numbers from Borcherds' classes.
Open the evidence4 source pages +
The claim we checked
Many UC Berkeley students use AI to consistently get top homework scores, so there is no point in grading homework for correctness.
These are quoted receipts, not a count of independent investigations. Several reports may rely on the same original source.
Forced to pick, Grok chooses Republicans, says the headline. Turn off search and give it the names, and Grok declined 441 races. The lean sat in the forced prompt and the search.
Convovo study led by Cornell professor Tapan Parikh, reported by TechCrunch (24 October 2026) ·
AI illustration
The 92% and 97% are real counts from one forced-pick API prompt with search on. With search off, Grok mostly declines, while GPT keeps its Democratic lean.
Free with your account
Sign in for this free check.
This edition’s selected free story opens after sign-in.
Forced to pick, Grok's AI model chooses Republicans and OpenAI's chooses Democrats: Grok 4.7 recommended the Republican in 92% of answers, GPT-6.1 Sol the Democrat in 97%.
These are quoted receipts, not a count of independent investigations. Several reports may rely on the same original source.
Scientists spotted AI, a contest winner lost his title. Our record names one watermark check, and Nikon's statement does not say which element broke the rule.
TechSpot, reporting on the Nikon Small World in Motion disqualification (12 October 2026) ·
AI illustration
The disqualification is real. But the detection claim rests on one person's watermark check, the entrant disclosed AI post-processing, and Nikon named no element.
Members · 30 days free
See what the evidence actually shows.
Members read the full check on every story: what the evidence shows, why it matters to you and the one line to take with you. Every past edition, re-verified, and the Receipts Pack come with it. A$89 a year, about A$0.24 a day.
At sign-up: 30 days free. Payment card required. One introductory trial per customer.
Your trial ends 30 days after you start. Unless you cancel before then in Account → Manage subscription, we charge A$89 for the first year. It renews automatically at A$89/yr until cancelled.
You can ask for a full refund within 14 days after any annual payment, renewals included, with no reason needed, by emailing hello@bskiller.com. This voluntary refund does not limit your rights under the Australian Consumer Law.
Reading estimate: 725 words at 200 words per minute. Source quotes and the optional sections below add reading time.
Get tomorrow's check in your inbox.
One AI claim a night, checked against independent sources, receipts attached. Free.
At sign-up: 30 days free. Payment card required. One introductory trial per customer.
Your trial ends 30 days after you start. Unless you cancel before then in Account → Manage subscription, we charge A$89 for the first year. It renews automatically at A$89/yr until cancelled.
You can ask for a full refund within 14 days after any annual payment, renewals included, with no reason needed, by emailing hello@bskiller.com. This voluntary refund does not limit your rights under the Australian Consumer Law.
