Prove your writing is human.
AI detectors guess. Humanums collects proof while you write — keystroke cadence, revision patterns, pause timing — and turns it into a verifiable badge anyone can inspect.
The case for slow writing
There is a quiet rebellion happening among people who still write for a living. It does not look like much from the outside — a cursor blinking, a coffee going cold, a paragraph written and rewritten six times before anyone else ever sees it.
The machines write faster. That is the whole point. Speed was the one thing humans could never win on, so it stopped being the thing worth winning. What is left is the thing a machine cannot fake — the record of a mind changing itself in public
Natural variance
14 thinking pauses
23 revisions, 8 reverts
No suspect pastes
3 sessions over 2 days
Trusted by independent writers, small publishers, and teams who sell work under their own name
AI detectors have a 61% false positive rate for non-native English speakers.
The internet is drowning in AI-generated content. By 2028, an estimated 90% of online content will be generated. And the tools meant to detect it — GPTZero, Originality.ai, Turnitin — are fundamentally unreliable.
They analyze finished text and guess. They flag human writers as AI. They miss AI text that's been lightly edited. Humanums takes the opposite approach.
Humanums doesn't detect AI. We certify humans.
Write. Certify. Prove.
Humanums records how a document is written, certifies the finished version, and generates a badge plus verification page that anyone can inspect.
Write
Draft inside our editor. As you type, the system silently records behavioral signals — keystroke cadence, pause distribution, revision patterns. Never content.
We never capture what you type — only how.
Certify
Click Certify. Our engine runs your writing through six behavioral signals that separate human composition from generated output, and produces a signed score.
Badge
Receive a verifiable badge and a public verification page. Embed it on your blog, portfolio, or contract deliverable. Anyone can click to inspect the proof.
What the process looks like on the wire.
Two paragraphs. Same text on the page. Radically different behavioral fingerprints.
One paste. Zero thinking.
Zero behavioral variance. No thinking pauses. No revisions. Pasted in 2 seconds.
47 minutes. 23 revisions. Real.
Natural timing variance. 14 thinking pauses. 23 revisions. Written across two sessions.
Behavioral proof beats probability guesses.
Detectors inspect output and tell you how likely it looks like a machine. Humanums captures the only thing a machine can't fake — the record of your mind changing itself in public.
Process, not output
Detectors inspect a finished artifact and guess. Humanums inspects the process. You can't retro-fit 47 minutes of natural composition.
Private by construction
We record timing and patterns, never keystrokes or content. The telemetry stream cannot reconstruct a single word you wrote.
Cryptographically signed
Every certificate carries a SHA-256 content hash and HMAC signature. Tamper-resistant verification any third party can check.
Built for the AI era
EU AI Act, California's AI Transparency Act, FTC guidance — regulation is pushing toward provenance, and Humanums is already there.
A mark readers can click, not a score they have to trust.
Every certified document gets a signed badge, a public verification page, and a tamper-resistant content hash. No account needed to verify. No black box.
Built for high-trust writing.
Humanums is most useful anywhere authorship matters — economically, reputationally, or academically.
Defend premium rates with proof
Attach a verification link to every deliverable and stop negotiating originality screenshot by screenshot.
Read the use caseAdd a trust layer to every byline
Contributor workflows, ghostwriter disclosure, and a visible proof layer readers can actually click on.
Read the use caseCarry evidence, not detector scores
Build behavioral proof while you write so you have something stronger than a probability score if your work is challenged.
Read the use caseStart free. Upgrade when you're ready.
Simple plans. No hidden fees. 14-day free trial of Creator Pro included.
Free
For writers testing the workflow.
- 3 certifications per month
- Badge with “Powered by” branding
- Public verification pages
- Basic portfolio page
- Email support
Creator Pro
For working writers who certify every piece.
- Unlimited certifications
- Clean badge without branding
- Custom portfolio URL
- HTML / Markdown / React embed codes
- Priority email support
Publisher
For teams protecting editorial trust.
- Everything in Creator Pro
- Up to 5 team members
- Custom branded badges
- Bulk certification
- API access
Humanums offers a free tier with 3 certifications per month. Creator Pro costs $9/month for unlimited certifications, and Publisher plans start at $29/month with team features and API access.
Payments processed securely by Paddle. Full refund within 14 days, no questions asked.
Learn where Humanums fits.
Deeper answers on certification vs. detection, the compliance landscape, and how writers are using behavioral proof today.
Humanums vs. AI detectors
Why probability scores create doubt, and why process-based proof creates something you can defend.
How Humanums works in practice
The draft, certification, badge, and verification flow — end to end.
Human authorship verification
The difference between proving authorship and merely screening for suspicious output.
Content authenticity badge
What makes a trust badge credible, and why the verification page behind it matters.
AI transparency compliance
How provenance and human-authorship proof fit broader transparency expectations.
Why Humanums exists
The product philosophy behind certification, provenance, and portable trust signals.
Answers before you sign up.
How is Humanums different from GPTZero or other AI detectors?
AI detectors analyze finished text and guess. They have high false positive rates and can be fooled. Humanums captures behavioral proof during writing. We don't detect AI — we certify humans.
What if I write in Google Docs and paste into your editor?
The telemetry will reflect a large paste event with no writing behavior. The system certifies content written in our editor today, and our Chrome extension for Google Docs is currently under review.
What data does Humanums collect?
Behavioral metadata only: keystroke timing (not the keys), pause durations, revision patterns, cursor movements, and paste sizes. We never capture content of keystrokes or pastes.
Can someone game the system by typing AI text manually?
It would require typing the entire document at natural pace with natural pauses and revisions. For 2,000 words, that's 60+ minutes of real work. Future versions will add style consistency analysis.
What happens after I certify a document?
You get a signed certificate, an embeddable badge, and a public verification page. The document becomes read-only and the content hash ensures it can't be altered.
Is my content private?
Yes. Your content is only accessible to you. The verification page shows writing statistics but never your content. The content hash allows verification without exposing text.
Do I need to write everything in one sitting?
No. Write across as many sessions as you want. Multi-session writing actually scores higher because it reflects natural human workflow.
In a world flooded with AI content, prove yours is real.
Create your first certificate in under five minutes. Free, no card required.