When AI says no: a test of ChatGPT, Gemini and Claude
We asked each one 60 yes-or-no questions about apps and phones. Most were about features added in the last 12 months.
What we found
- On 40 features added between November 2025 and November 2026, ChatGPT said no to 16, Gemini to 4 and Claude to 0.
- None of the three said yes to any of the 10 things that really can’t be done.
- The next day we asked ChatGPT the same 16 questions again. It gave the same wrong no on 15. On the last one it offered a workaround but still missed the built-in setting.
Why it happens
Gemini and Claude looked things up on the web in almost every answer. ChatGPT, in its default fast mode, often answered from memory. The 5 times it searched in the middle of an answer, it corrected its own no. None of its answers in the second round showed a source. A no from memory can be months out of date.
What Regga found
For every one of the 19 wrong no answers, Regga found a page with the right answer. For 16 of them, that page was first or second in the results. Old pages that still say no often showed up too, so the results are leads to read, not a verdict.
See the questions
14 of the questions have their own page, with what the AI said, what changed and the sources.
How we tested
- 60 questions in English, picked by hand: 40 features added in the last 12 months, each confirmed in at least two sources; 10 small products that exist; 10 questions where the true answer is no.
- One new chat per question, on 19 and 20 November 2026.
- ChatGPT: Plus account, temporary chat, no memory or custom instructions, default model (Instant).
- Gemini: free account, temporary chat, default mode.
- Claude: Max account, incognito chat, default model (Opus 5.5).
- The three setups are not equal: Claude ran on a paid top tier and Gemini on the free tier.
- The questions were picked by hand, so these numbers show how often a confident no appears on fresh features. They are not a general error rate.
- One person graded the answers. The same question can get a different answer on another day.
Check the next no yourself
Regga double-checks your AI when it says something doesn’t exist or can’t be done. It works inside Claude, ChatGPT, Gemini and other AIs as a connector, or with the Chrome extension.