Hide names from artificial intelligence.
100% freeNo sign-upsNo trackingKeeps working with the internet off
Use ChatGPT, Gemini, Claude, Copilot, Meta Artificial Intelligence, Grok, DeepSeek, Qwen, any other artificial intelligence, or an app that offers several artificial intelligences in one place, on a message about real people, without handing over their details. Swap the names and numbers for labels, ask the artificial intelligence, then put the real details back into its reply. Everything happens on this phone or computer.
The short answer. Yes, you can use any artificial intelligence on a message about real people without handing over their details. Paste the message here, swap the names and numbers for labels, ask the artificial intelligence, then paste its reply back to put the real details in again. Everything happens on your own phone or computer.
Watch it work. A 39-second demonstration on YouTube, recorded with the internet switched off.
Paste the message
This can also catch an organisation or a place. You can untick those in the list that follows.
First names on their own are hidden too, such as Marisol, Wanjiru, Lakshmi or Kamau, using a list of common first names from many countries that sits inside this page. Streets and places, such as Moi Avenue, are left as they are. Your text stays on your own phone or device the whole time. Please still read the result before you paste it into an artificial intelligence, because a name that is not on the list can be missed.
Read it, then copy it into your artificial intelligence
The yellow labels stand in for the real details. Read it through before you copy it. Job titles, employers and city names are left in unless you add them to “Also hide these words” above, and a first name that is not on the built-in list of common first names, or a first name that is also an ordinary word and begins a sentence, such as Rose or Will, can still be missed. If anything was missed, add it there and press Swap again.
What was swapped
Untick anything that did not need hiding. “Sure” means the detail has a fixed shape the tool can test, like a card number or an email address. “Check this” means it was worked out from clues in the message, like a name after “Hi”, so make sure it is right and nothing is missing.
Paste the artificial intelligence’s reply
The labels in the reply are turned back into the real details, here on this device. Keep this page open until you have finished: if you close it, the list of swaps is gone.
Your finished reply
What it will not do
It finds names by the clues around them: “Hi Sarah”, “My name is”, “Mr”, a signature, a WhatsApp sender, or a name written next to a phone number, an email or an address. Careful mode is ticked unless you untick it. It hides the name that follows a title (the title itself, such as Councillor or Doctor, stays in the message), and it treats two or more capitalised words in a row as a name. That can catch a place or an organisation as well, and you can untick those in the list below. Job titles, roles and city names are left in unless you add them to “Also hide these words”. A name typed in small letters with nothing around it can still be missed, so always read the swapped message, and add anything it missed yourself.
The artificial intelligence still sees the rest of the message: the complaint, the product and the amounts. And this is not a legal check of any kind. It does one thing: the details you swap never leave this device.
What you paste here stays on this device and is not saved. Turn the internet off and it still works.
How well it works
It swaps the names, numbers and addresses it recognises in WhatsApps, emails and complaints written the way people write in South Africa, Nigeria, Kenya, India, the United Kingdom and the United States and beyond. It may also swap something that did not need hiding, such as a company’s own email address, so read the swapped text before you send it.
Which numbers and countries it recognises
It recognises phone numbers written the way they are written in the United States, Canada, the Caribbean, the United Kingdom, Ireland, Australia, New Zealand, South Africa, Nigeria, Kenya, Ghana, Uganda, Zambia, Zimbabwe, Botswana, India, Singapore, Malaysia, the Philippines and across Europe, with or without the country code. It knows the shape of number plates in the United Kingdom, Ireland, South Africa, Kenya, Nigeria, Ghana, Uganda, Tanzania, Botswana, India, France, Germany and the Netherlands, and any plate anywhere when the message calls it a plate, a registration or a rego. It knows identity and health numbers including United Kingdom National Insurance and the National Health Service numbers, United States Social Security and Medicare numbers, Canadian social insurance and health card numbers, Irish PPS numbers, Australian Medicare and tax file numbers, New Zealand NHI and IRD numbers, South African identity numbers, Nigerian NIN and BVN, the Ghana Card, Kenyan and Zimbabwean identity numbers, and Indian Aadhaar and PAN numbers. We test every one of these with made-up examples, and with sentences that have no personal details in them, so ordinary numbers such as prices, dates and population figures are left alone.
Then we wrote 100 harder ones: names typed in small letters, names that are also ordinary words like Grace and Precious, surnames written first, and addresses in Lagos, Nairobi, Accra, Soweto and Mumbai. On the first try it found only 34 of the 80 details. We changed it until it found all 80. With Careful mode unticked it leaves messages with no personal details alone; with it ticked it can also hide an organisation or a place, which you can untick in the list.
That is a small test, and we improved the tool on those same messages, so real messages will go less well. That is why we keep testing it on new messages it has never seen, and why you should always read the swapped message before you copy it.
Why this exists
In a March 2026 survey of 492 knowledge workers and 292 executives in seven countries, 52% of the workers said they used artificial intelligence tools their employer had not approved, while 90% of the executives said they were confident they could see which tools staff used.
In a 2025 study of 48,000 people in 47 countries, including South Africa, Nigeria and Egypt, almost half of workers admitted using artificial intelligence against their employer’s rules, including putting customer information into free public artificial intelligence tools.
Sources: Okta survey, reported by The Register, May 2026 (Okta sells workplace security tools) · KPMG and the University of Melbourne, 2025.