AI Content Tell Checker
Paste a draft and see the words, phrases, and patterns that make it read AI-written, highlighted in place. Not an AI detector: an editor's checklist. Runs entirely in your browser.
What are AI tells?
AI tells are stylistic patterns common in unedited AI drafts: overused words like "delve" and "tapestry", formulaic openers, heavy hedging, uniform sentence rhythm, and em dash overuse. No single tell proves AI authorship, but together they make content read generic, and generic underperforms.
Plain text works best. 50+ words minimum so the rhythm checks mean something.
Runs locally in your browser. Nothing is uploaded or stored.
Human-edit score
- AI-overused phrases
- Hedge words
- Passive voice
- Formulaic openers
- Em dashes
- Sentence rhythm
Your text, highlighted
What to fix
Don't want to fix it yourself?
Removing tells is exactly what my AI Content Humanization service does: line-by-line rewriting into your brand voice, fact-checking included, from $0.09/word.
AI tells: frequently asked questions
No. This tool cannot determine whether a human or an AI wrote a text, and no tool can do that reliably. It flags stylistic patterns that overlap with typical unedited AI output, so you can edit them out. Human writing can trip every flag; well-edited AI writing can pass clean.
AI tells are stylistic patterns common in unedited AI drafts: overused vocabulary like "delve" and "tapestry", formulaic openers like "In today's fast-paced world", heavy hedging, uniform sentence lengths, and em dash overuse. None proves AI authorship; together they make text read generic.
Google does not penalize AI-assisted content as such; it rewards helpful, original content and demotes thin, generic content. AI tells correlate with the generic style that underperforms, which is why editing them out matters.
Removing tells improves readability and trust, but ranking also needs entity coverage, search intent alignment, internal linking, and real expertise. Style is one layer of a content strategy, not the whole thing.
Disclaimer
This is an experimental, educational tool provided as-is, with no warranty. It is NOT an AI detector: it cannot determine who or what wrote a text, and it must never be used to accuse anyone of using AI, grade student work, or make hiring or publication decisions. It flags stylistic patterns using a transparent, rule-based word and pattern list; human writing regularly trips these flags, and well-edited AI text passes them. English only. Everything you paste stays in your browser; nothing is uploaded or stored. The score reflects style, not authorship and not ranking potential.
Publishing AI content at scale?
The checker finds the tells. My humanization service removes them, verifies the facts, and keeps your voice. Start with a free sample edit.
Reply guaranteed within 24 hours.