Content at Scale AI detector: how accurate, and where it is now
Short answer Content at Scale became BrandWell in August 2024, and its free detector now runs as a Moxby browser add-on that says it can't tell who wrote a text. Its old 98% accuracy claim didn't hold: the FTC cited about 53% on non-academic text. Treat any score as an editing hint, never proof.
Want help applying this to your business? Talk it through on a free 30-minute call.
Book a callWhere is the Content at Scale AI detector now?
It changed hands twice. Content at Scale renamed itself BrandWell on August 27, 2024 (PRWeb). Today the old detector address, contentatscale.ai/ai-content-detector, redirects to BrandWell, which redirects again to a listing on Moxby, which BrandWell's home page calls "our sister product" (BrandWell).
| Version | Where | Cost | What it claims |
|---|---|---|---|
| Free detector | Moxby browser add-on ("Mod") | Free plan (Moxby) | Reviews up to 10,000 characters for repetitive patterns; says it doesn't decide who wrote the text (Moxby Marketplace) |
| Inside RankWell | BrandWell's writing and SEO tool, billed as "the original polished SEO & AI Detector" | $249, $499 or $999 a month (BrandWell Pricing) | "Detection and review tools" for editors; no accuracy figure (BrandWell RankWell) |
What's gone: the old free web box and the 98% claim. BrandWell's pricing page lists no standalone detector plan and no accuracy figure today (BrandWell Pricing). To use the free version you install the Moxby browser extension, and your text goes "through the Moxby gateway to the analysis service" (Moxby Marketplace). Don't paste client or confidential text.
Is the Content at Scale AI detector accurate?
No independent test backs it. Two public records say the opposite.
The FTC case. The FTC's filing names the company "Workado, LLC, f/k/a Content at Scale AI" (f/k/a means "formerly known as") and summarizes its complaint:
"testing shows the AI model powering Workado's detector could correctly classify non-academic AI-generated text with around 53 percent accuracy." (Federal Register, May 2, 2025)
The company had claimed "over 98 percent accuracy" on text from ChatGPT, Claude and GPT-4. The complaint says the model was trained "to accurately classify only academic content, rather than the types of marketing content submitted by Workado users" (Federal Register). In other words, a tool sold to marketers was tuned on school papers. The FTC approved a final order on August 28, 2025: no accuracy claims without "competent and reliable evidence," and customers had to be emailed about the settlement (FTC).
The university test. In 2023, eight researchers ran 14 detectors on 54 documents. Content at Scale labeled every human-written text correctly, and every AI-written one as human too: 33% overall, last of 14. The authors called the tool "inherently biased towards human classification and thus completely useless" (Weber-Wulff et al., 2023).
Its bias explains the old reviews praising it: a tool that says "human" to everything never falsely accuses anyone, and never catches anything.
What does the detector say about itself now?
The Moxby listing is blunt:
"Results are approximate and can be wrong. They do not verify authorship and should not be used as proof in academic, employment, legal, or disciplinary decisions." (Moxby Marketplace)
It also warns that short, technical or templated human writing gets flagged. Many service pages, like a price list or a step-by-step booking guide, are exactly that. The tool now calls its result "guidance" for finding flat, repetitive passages, which is the only honest use.
Should a business owner use it?
Only as an editing hint. A detector score doesn't decide your search rankings; Google judges the page, not who typed it. See AI detectors and SEO and Does Google penalize AI content?
| Your goal | Does this detector help? | Better check |
|---|---|---|
| Rank on Google | No | Facts a person checked, your own photos, prices and experience (quality framework) |
| Check a freelance writer | No: it can't prove who wrote it | Ask for drafts and notes; check facts and anything only your business would know |
| Find flat, robotic paragraphs | Somewhat | Read it aloud; cut filler; add one specific detail per section |
| Prove your own writing is human | No | Your drafts, version history and first-hand details |
Example
A roofer checking a hired writer
The owner pays $300 for five service pages. The free detector flags two as "robotic." That alone proves nothing. He checks the pages instead: one lists a shingle warranty he doesn't offer. He sends it back for that, not for the score.
An accountant pasting client notes
She drafts a newsletter from client emails and runs it through the add-on. The text leaves her computer for an outside service. She deletes names and numbers first, or skips the check.
A cleaning company on RankWell
The team pays $249 a month for RankWell's writing and SEO tools. The detector is one input. Pages still need a person to fix wrong facts and add real job details before they go live.
FAQ
Is the Content at Scale AI detector still free?
Yes, in its new form: a free Moxby browser add-on that reviews up to 10,000 characters at a time (Moxby Marketplace). You need the Moxby extension installed.
Is BrandWell the same company as Content at Scale?
Yes. Content at Scale announced its rename to BrandWell on August 27, 2024 (PRWeb).
Did the FTC fine Content at Scale?
The FTC's releases describe an order, not a fine: no unsupported accuracy claims, keep the evidence, email customers and report on compliance (FTC).
Is another AI detector more reliable?
The same 2023 study found none of 14 tools accurate or reliable, and none caught every AI text after editing or paraphrasing (Weber-Wulff et al., 2023). Switching tools doesn't fix the problem.
Different for you?
- Worried a detector score hurts your rankings: AI detectors and SEO
- Publishing AI-drafted pages: Does Google penalize AI content?
- Your pages lack proof you know the work: E-E-A-T for a small business
- You want pages written and checked for you: SEO agent