# One time pad ciphers in emails

In order to achieve very high security for privacy, would it be cryptographically secure to use one time pad ciphers in emails? The distribution of the keyword would pose no problems since I would distribute it to other people in person, therefore avoiding interception.

Just to avoid confusion, by keyword I mean a random string of bits as long as the message.

• It would certainly offer no practical advantage over common methods such as OpenPGP. Whether it would be feasible basically boils down to how much true randomness you can generate, which is something I am not very familiar with. Note that if you are using a pseudorandom generator, it's no longer a one-time pad. – fkraiem Jul 31 '15 at 21:12
• The randomness is generated by observing and recording sounds from radiation with a gigameter. Radiation is a truly random thing. I would then compare the sounds of the radiation to the alphabet in order to generate a truly random one time pad. – Anonymous Jul 31 '15 at 21:16
• Well, if you are satisfied that you can generate a sufficient amount of true randomness, it is trivial to write a program which XORs some randomness with any file of your choice, and attach the result in an email. – fkraiem Jul 31 '15 at 21:18
• More random than what? – fkraiem Jul 31 '15 at 21:27
• …would it be smart… is practically fishing for personal opinions (which is off-topic at Crypto.SE). Maybe you could reformulate that into …would it be cryptographically secure… or something alike? (that would make the question on-topic) – e-sushi Jul 31 '15 at 23:54

However, OTP alone offers no authentication and leaves the message malleable. If Alice sends a message Y to Bob, standing for 'yes', Mallory can guess this and modify the message to say N instead (but he would garble the message if he thought it said something different).