MATHS BOFFIN Adi Shamir, a professor at the Weizmann Institute of Science in Israel, has done some sums and come up with the conclusion that a single slip-up chips could blow public key cryptography wide open to hackers.

Shamir has written a paper, which has been seen by only few of his fellow maths boffins, in which he claims that a maths error as yet unknown to but potentially discovered by a bad guy could lead to catastrophic consequences.

Executing the attack would require only knowledge of the maths flaw and the ability to send a "poisoned" encrypted message to a protected computer, he wrote. It would then be possible to compute the value of the secret key used by the targeted system.

It might be a good idea to listen to Shamir. Since the A in RSA stands for (Adi) Shamir, after he came up with the idea with MIT buddies Ron Rivest and Leonard Adleman.

It seems that once someone exploits this hypothetical error, millions of PCs could be attacked simultaneously, without having to manipulate the operating environment of each one of them individually.

This might not have been a problem if Shamir's paper had not been spotted by John Markoff of The New York Times. Shamir told Markoff that he is unaware of anyone exploiting such a vulnerability. But they might, he warned. Cryptically. µ