How I used AI LLMs as an effective Null Cipherer to hide a message in plain sight.
What if you could hide a message in plain sight and no one would ever notice?
How I used AI LLMs as an effective Null Cipherer to hide a message in plain sight.

What if you could hide a message in plain sight and no one would ever notice?
Context: You are in a confined area and all eyes are on you. You can communicate freely but you are being watched. All messages going in and out of your area are read by a gatekeeper. He gets to decide if your letter reaches your mother/friend/colleague.
You need to send a message out… But what do you do? One method that we could use is Null Cipher. This technique hides a message within a message. Say for example we want to hide the sentence ‘start compiling the code now’, the message would be:
Many travellers often eargly start exploring the markets while compiling recipes that truly define the essence of a culinary code respected by every chef now.
This way, a typical reader will read the sentence as-is, but for the targetted reader, they would know what to look for and extract the hidden message out of it. For such a simple message, could take hours to craft, especially for the typical layperson. Therefore, we can use LLMs, to build such a message.
The prompt would be something like this:
‘I want to hide a message using Null Cipher, can you create a letter to describe your holiday at Vietnam, while inserting ther words ‘start compiling the code now’ in every 5th word of the paragraph. The rules are, the first letter can start anywhere, the ending does not have to be the last word, generate 10 possible sentences and the words have to be grammatical and sound logical. You cannot just insert the words in-between the sentences. It has to flow with the whole sentence and sound right.’
Put this into Gemini or ChatGPT to get your answer!
- disclaimer, you often have to tune your prompt to get good answers. Really depends on the mood of the LLMs and weather to give you a good answer. Keep trying until you get the right message that you want.
If you like this post, please give a thumbs up. In my next article, i will share how we can use LLMs to extract hidden messages.
메타데이터
- post_id
- 3bcc303e256f
- slug
- using-llms-as-an-effective-null-cipherer-3bcc303e256f
- url
- https://medium.com/@tmnet/using-llms-as-an-effective-null-cipherer-3bcc303e256f
- canonical_url
- https://medium.com/@tmnet/using-llms-as-an-effective-null-cipherer-3bcc303e256f
- author_url
- https://medium.com/@tmnet
- status
- ok
- fetched_at
- 2026-07-10 20:57:38