Its developers claim this will protect senders from having ill-judged electronic words used against them later.

The most high-profile instance came last year when Microsoft's Bill Gates had to defend himself against his own e-mails in a US antitrust case.

Hack it if you can

UK company Global Markets Research (GMR) designed 1on1 e-mail "to guarantee complete confidentiality".

It uses 2,048-bit public key encryption to secure the message in transit and GMR have such confidence in it that they are offering $50,000 to anyone who can hack into a message within three months.

The self-destruct feature is called autoshredder and the package also prevents recipients from just cutting and pasting out of it. "That would be pointless," GMR's technical director, Steven James told New Scientist magazine.

1on1mail also ensures that the e-mail is not stored anywhere on the recipient's computer. Finally, when the message self-destructs, it is overwritten on the disk, so it cannot be undeleted later.

Gimmick jibe

However, critics have been quick to give their views. "2,048-bit encryption is ridiculous," cryptographer Bruce Schneier told technical news Website ZDNN. "It is irrelevant. The security is determined by the password anyway. If the user picks a bad one, the security is bad."

Hushmail, a rival encrypted e-mail service, dismissed the self-destruct feature as a gimmick.