Hand Typed
By
You
Hand Typed
By
You
When you care about others, you don't put junk in their inbox. Send them hand-typed emails or WhatsApp messages. Let your audience know a human is writing to them.
When you care about others, you don't put junk in their inbox. Send them hand-typed emails or WhatsApp messages. Let your audience know a human is writing to them.
When you care about others, you don't put junk in their inbox. Send them hand-typed emails or WhatsApp messages. Let your audience know a human is writing to them.
It's Me
There may not be a single switch, but there are clear steps forward.
Every path is different. These are the ways we help people move forward with confidence.
Showing you typed something yourself can be as simple as flipping a switch
You’re closer than you think
Real Humans
Real Humans
At ActuallyMe, we think the effort of writing something yourself deserves to be seen. Every keystroke, signed and timestamped, so what you send carries proof of the person who made it
At ActuallyMe, we think the effort of writing something yourself deserves to be seen. Every keystroke, signed and timestamped, so what you send carries proof of the person who made it
At ActuallyMe, we think the effort of writing something yourself deserves to be seen. Every keystroke, signed and timestamped, so what you send carries proof of the person who made it
How It Works
How It Works
Typing a message yourself is a small act of care. Our tool allows you to make it visible, without getting in the way. Here is how it fits into the way you already write.
Type it
Open our app and start typing. The text field works the way you expect: clean, quiet, subtle. Every keystroke is recorded, so the rhythm of your thinking is preserved. No pasting allowed, no shortcuts. Just you and the words you chose.
Type it
Open our app and start typing. The text field works the way you expect: clean, quiet, subtle. Every keystroke is recorded, so the rhythm of your thinking is preserved. No pasting allowed, no shortcuts. Just you and the words you chose.
Send it or post it
Copy your message and send it through whatever you already use: email, WhatsApp, LinkedIn, whoever your reader is. A short link will accompany your message. Your message stays yours. Your message is encrypted on your device before it is saved. We store the proof, not the words. Even we cannot read what you wrote
Send it or post it
Copy your message and send it through whatever you already use: email, WhatsApp, LinkedIn, whoever your reader is. A short link will accompany your message. Your message stays yours. Your message is encrypted on your device before it is saved. We store the proof, not the words. Even we cannot read what you wrote
Let them verify
Your recipient clicks the link and sees a simple page: a green check, and a line that confirms it was written by hand. They do not need an account or an app. It takes a few seconds, and it tells them something most inboxes have forgotten how to say: A real person sat down and wrote to them.
Let them verify
Your recipient clicks the link and sees a simple page: a green check, and a line that confirms it was written by hand. They do not need an account or an app. It takes a few seconds, and it tells them something most inboxes have forgotten how to say: A real person sat down and wrote to them.
Your questions.
Answered.
The obvious ones, the skeptical ones, and the ones your IT department will ask.
Still got questions? We'd rather explain than assume. Drop us a line.
Can't someone just fake the typing patterns?
Can't someone just fake the typing patterns?
In theory, yes. In practice, almost no one does. Faking the rhythm of human writing means reading AI output off a second screen and typing it out by hand, complete with pauses and corrections. By the third message, most people realise it is faster to just write the thing themselves. That is the point. We are not trying to be uncrackable, we are trying to make care the path of least resistance.
Do you read what I write?
Do you read what I write?
No. Your message is encrypted on your device before anything leaves your machine. The key lives in the link you share with your recipient, not on our servers. We hold the proof that typing happened, not the words themselves. Even if we wanted to read your message, we could not. Our client code is open source so anyone can verify this for themselves.
How is this different from GPTZero or Turnitin?
How is this different from GPTZero or Turnitin?
Those tools guess whether text was written by an AI, after the fact, by analysing the words. They are often wrong, in both directions. We do something different: we record how the text was made, as it is being made. No guessing, no probability scores, no false accusations. Either a message was typed by hand through Humanity Keys, or it was not. It is proof of process, not a verdict on the output.
Your questions.
Answered.
The obvious ones, the skeptical ones, and the ones your IT department will ask.
Still got questions? We'd rather explain than assume. Drop us a line.
Can't someone just fake the typing patterns?
Can't someone just fake the typing patterns?
In theory, yes. In practice, almost no one does. Faking the rhythm of human writing means reading AI output off a second screen and typing it out by hand, complete with pauses and corrections. By the third message, most people realise it is faster to just write the thing themselves. That is the point. We are not trying to be uncrackable, we are trying to make care the path of least resistance.
Do you read what I write?
Do you read what I write?
No. Your message is encrypted on your device before anything leaves your machine. The key lives in the link you share with your recipient, not on our servers. We hold the proof that typing happened, not the words themselves. Even if we wanted to read your message, we could not. Our client code is open source so anyone can verify this for themselves.
How is this different from GPTZero or Turnitin?
How is this different from GPTZero or Turnitin?
Those tools guess whether text was written by an AI, after the fact, by analysing the words. They are often wrong, in both directions. We do something different: we record how the text was made, as it is being made. No guessing, no probability scores, no false accusations. Either a message was typed by hand through Humanity Keys, or it was not. It is proof of process, not a verdict on the output.
Your questions.
Answered.
The obvious ones, the skeptical ones, and the ones your IT department will ask.
Can't someone just fake the typing patterns?
Can't someone just fake the typing patterns?
In theory, yes. In practice, almost no one does. Faking the rhythm of human writing means reading AI output off a second screen and typing it out by hand, complete with pauses and corrections. By the third message, most people realise it is faster to just write the thing themselves. That is the point. We are not trying to be uncrackable, we are trying to make care the path of least resistance.
Do you read what I write?
Do you read what I write?
No. Your message is encrypted on your device before anything leaves your machine. The key lives in the link you share with your recipient, not on our servers. We hold the proof that typing happened, not the words themselves. Even if we wanted to read your message, we could not. Our client code is open source so anyone can verify this for themselves.
How is this different from GPTZero or Turnitin?
How is this different from GPTZero or Turnitin?
Those tools guess whether text was written by an AI, after the fact, by analysing the words. They are often wrong, in both directions. We do something different: we record how the text was made, as it is being made. No guessing, no probability scores, no false accusations. Either a message was typed by hand through Humanity Keys, or it was not. It is proof of process, not a verdict on the output.
Still got questions? We'd rather explain than assume. Drop us a line.
Your questions.
Answered.
New to all this? These answers cover how it works, what's signed, and what the proof actually means.
Still have a question? Send us a message, we'll get back to you fast.
How do I know if ActuallyMe is right for me?
If you write things that people need to trust, articles, legal documents, exam answers, client correspondence, then being able to show a verified person typed them, by hand, in one sitting, is worth having. If your work could be quietly replaced with pasted or generated text and you'd have no way to show otherwise, this is for you.
How do I know if ActuallyMe is right for me?
If you write things that people need to trust, articles, legal documents, exam answers, client correspondence, then being able to show a verified person typed them, by hand, in one sitting, is worth having. If your work could be quietly replaced with pasted or generated text and you'd have no way to show otherwise, this is for you.
What exactly gets signed?
What exactly gets signed?
Every keystroke, as you type it. We cryptographically sign the content, the timing, and your verified identity, then bundle it into a certificate. The result is a tamper-evident record that a specific verified person typed specific words at a specific moment.
Does this prove a human wrote it, or that AI didn't?
Does this prove a human wrote it, or that AI didn't?
It proves a real, verified person physically typed the content, by hand, rather than pasting or auto-generating it. What it does not do is judge where the ideas came from. We're deliberate about this: actually.me certifies manual entry by a verified person, not original authorship. That alone is something pasted or generated text cannot show, and for most uses it's exactly what matters.
Could someone just copy text from an AI and type it out by hand?
Could someone just copy text from an AI and type it out by hand?
In principle, yes, and we'd rather say so plainly than pretend otherwise. ActuallyMe proves the typing was manual and tied to a verified person, not that the ideas originated in their head. We add signals that make transcription harder to pass off as composition, but we don't claim to read minds. If a use case needs more than verified manual entry, we'll tell you straight.
Do I need special hardware, or does it work on my current devices?
Do I need special hardware, or does it work on my current devices?
Right now, ActuallyMe runs as software on the devices you already use, and gets you a verified link straight away. We're also developing a dedicated keyboard, with a secure enclave and fingerprint scanner, for the higher-assurance use cases like notary, legal, and compliance work where the record needs to hold up under the closest scrutiny. The hardware is in active development, and we'll share more as it gets closer.
What does the person receiving my work actually see?
What does the person receiving my work actually see?
A link. They click it and see a certificate: a replay of the writing as it happened, a timestamp, and confirmation the content matches and was typed by your verified identity. No technical knowledge needed, the same way nobody needs to understand encryption to trust a DocuSign envelope.
Is my biometric data safe?
Is my biometric data safe?
For the software, verification ties to your account identity. For the keyboard we're building, the plan is for your fingerprint to unlock the signing key locally on the device and never be uploaded or stored by us. The principle is the same either way: we hold the proof, not your biometrics.
Can the certificate be tampered with after the fact?
Can the certificate be tampered with after the fact?
The certificate records exactly what was typed. When you receive someone's work, you can check it against their ActuallyMe link, if even one character differs from what's on record, you'll see it doesn't match. Nobody can quietly alter the text and keep a valid certificate, because the altered version simply won't line up with the one that was signed.
Your questions.
Answered.
New to all this? These answers cover how it works, what's signed, and what the proof actually means.
Still have a question? Send us a message, we'll get back to you fast.
How do I know if ActuallyMe is right for me?
If you write things that people need to trust, articles, legal documents, exam answers, client correspondence, then being able to show a verified person typed them, by hand, in one sitting, is worth having. If your work could be quietly replaced with pasted or generated text and you'd have no way to show otherwise, this is for you.
How do I know if ActuallyMe is right for me?
If you write things that people need to trust, articles, legal documents, exam answers, client correspondence, then being able to show a verified person typed them, by hand, in one sitting, is worth having. If your work could be quietly replaced with pasted or generated text and you'd have no way to show otherwise, this is for you.
What exactly gets signed?
What exactly gets signed?
Every keystroke, as you type it. We cryptographically sign the content, the timing, and your verified identity, then bundle it into a certificate. The result is a tamper-evident record that a specific verified person typed specific words at a specific moment.
Does this prove a human wrote it, or that AI didn't?
Does this prove a human wrote it, or that AI didn't?
It proves a real, verified person physically typed the content, by hand, rather than pasting or auto-generating it. What it does not do is judge where the ideas came from. We're deliberate about this: actually.me certifies manual entry by a verified person, not original authorship. That alone is something pasted or generated text cannot show, and for most uses it's exactly what matters.
Could someone just copy text from an AI and type it out by hand?
Could someone just copy text from an AI and type it out by hand?
In principle, yes, and we'd rather say so plainly than pretend otherwise. ActuallyMe proves the typing was manual and tied to a verified person, not that the ideas originated in their head. We add signals that make transcription harder to pass off as composition, but we don't claim to read minds. If a use case needs more than verified manual entry, we'll tell you straight.
Do I need special hardware, or does it work on my current devices?
Do I need special hardware, or does it work on my current devices?
Right now, ActuallyMe runs as software on the devices you already use, and gets you a verified link straight away. We're also developing a dedicated keyboard, with a secure enclave and fingerprint scanner, for the higher-assurance use cases like notary, legal, and compliance work where the record needs to hold up under the closest scrutiny. The hardware is in active development, and we'll share more as it gets closer.
What does the person receiving my work actually see?
What does the person receiving my work actually see?
A link. They click it and see a certificate: a replay of the writing as it happened, a timestamp, and confirmation the content matches and was typed by your verified identity. No technical knowledge needed, the same way nobody needs to understand encryption to trust a DocuSign envelope.
Is my biometric data safe?
Is my biometric data safe?
For the software, verification ties to your account identity. For the keyboard we're building, the plan is for your fingerprint to unlock the signing key locally on the device and never be uploaded or stored by us. The principle is the same either way: we hold the proof, not your biometrics.
Can the certificate be tampered with after the fact?
Can the certificate be tampered with after the fact?
The certificate records exactly what was typed. When you receive someone's work, you can check it against their ActuallyMe link, if even one character differs from what's on record, you'll see it doesn't match. Nobody can quietly alter the text and keep a valid certificate, because the altered version simply won't line up with the one that was signed.
Your questions.
Answered.
New to all this? These answers cover how it works, what's signed, and what the proof actually means.
How do I know if ActuallyMe is right for me?
If you write things that people need to trust, articles, legal documents, exam answers, client correspondence, then being able to show a verified person typed them, by hand, in one sitting, is worth having. If your work could be quietly replaced with pasted or generated text and you'd have no way to show otherwise, this is for you.
How do I know if ActuallyMe is right for me?
If you write things that people need to trust, articles, legal documents, exam answers, client correspondence, then being able to show a verified person typed them, by hand, in one sitting, is worth having. If your work could be quietly replaced with pasted or generated text and you'd have no way to show otherwise, this is for you.
What exactly gets signed?
What exactly gets signed?
Every keystroke, as you type it. We cryptographically sign the content, the timing, and your verified identity, then bundle it into a certificate. The result is a tamper-evident record that a specific verified person typed specific words at a specific moment.
Does this prove a human wrote it, or that AI didn't?
Does this prove a human wrote it, or that AI didn't?
It proves a real, verified person physically typed the content, by hand, rather than pasting or auto-generating it. What it does not do is judge where the ideas came from. We're deliberate about this: actually.me certifies manual entry by a verified person, not original authorship. That alone is something pasted or generated text cannot show, and for most uses it's exactly what matters.
Could someone just copy text from an AI and type it out by hand?
Could someone just copy text from an AI and type it out by hand?
In principle, yes, and we'd rather say so plainly than pretend otherwise. ActuallyMe proves the typing was manual and tied to a verified person, not that the ideas originated in their head. We add signals that make transcription harder to pass off as composition, but we don't claim to read minds. If a use case needs more than verified manual entry, we'll tell you straight.
Do I need special hardware, or does it work on my current devices?
Do I need special hardware, or does it work on my current devices?
Right now, ActuallyMe runs as software on the devices you already use, and gets you a verified link straight away. We're also developing a dedicated keyboard, with a secure enclave and fingerprint scanner, for the higher-assurance use cases like notary, legal, and compliance work where the record needs to hold up under the closest scrutiny. The hardware is in active development, and we'll share more as it gets closer.
What does the person receiving my work actually see?
What does the person receiving my work actually see?
A link. They click it and see a certificate: a replay of the writing as it happened, a timestamp, and confirmation the content matches and was typed by your verified identity. No technical knowledge needed, the same way nobody needs to understand encryption to trust a DocuSign envelope.
Is my biometric data safe?
Is my biometric data safe?
For the software, verification ties to your account identity. For the keyboard we're building, the plan is for your fingerprint to unlock the signing key locally on the device and never be uploaded or stored by us. The principle is the same either way: we hold the proof, not your biometrics.
Can the certificate be tampered with after the fact?
Can the certificate be tampered with after the fact?
The certificate records exactly what was typed. When you receive someone's work, you can check it against their ActuallyMe link, if even one character differs from what's on record, you'll see it doesn't match. Nobody can quietly alter the text and keep a valid certificate, because the altered version simply won't line up with the one that was signed.
Still have a question? Send us a message, we'll get back to you fast.
Talk to us
Get early access
Maybe you're a novelist who's been accused of using ChatGPT. Maybe you run a law firm that needs airtight document provenance. Maybe you're a professor who knows GPTZero is wrong half the time but has nothing better to offer your students. Whatever brought you here, we want to talk.
Prefer to chat first? Send us an email, we’re always happy to help.
Talk to us
Get early access
Maybe you're a novelist who's been accused of using ChatGPT. Maybe you run a law firm that needs airtight document provenance. Maybe you're a professor who knows GPTZero is wrong half the time but has nothing better to offer your students. Whatever brought you here, we want to talk.
Prefer to chat first? Send us an email, we’re always happy to help.
Talk to us
Get early access
Maybe you're a novelist who's been accused of using ChatGPT. Maybe you run a law firm that needs airtight document provenance. Maybe you're a professor who knows GPTZero is wrong half the time but has nothing better to offer your students. Whatever brought you here, we want to talk.
Prefer to chat first? Send us an email, we’re always happy to help.