ONE OF THE MOST SECURE METHODS of transmitting data has been defeated by hackers armed with lasers.
Quantum cryptography is supposed to be really tough to defeat because, if a third party does intercept a quantum signal, the tampering becomes apparent to parties at both ends.
But quantum hackers at the Norwegian University of Science and Technology in Trondheim have performed successful hacks of two commercial quantum cryptographic systems and managed to do it without detection.
Quantum encryption is based on the principle that you cannot look at a quantum message without disturbing it. A message is encoded into a beam of light using two different quantum states of photons. The receiver has a detector that measures the quantum states of the incoming photons. But if you look at the signal in between you will change the message.
According to Popsci, the hackers got around the rules of quantum physics by simply intercepting the incoming signal and generating a brand new one to send on to the receiver.
They shone a continuous 1-milliwatt laser at the receiver's detector, blinding it while they intercepted the sender's signal.
While blinded, the detector functions as a classical light detector. The interceptor receives the sender's signal and pumps an extra bright pulse of light at the receivers' detector every time it reads a "one" in the original signal.
The receiver gets the correct signal from the interceptor even though it's forged, making it a classical physics signal rather than a quantum one, so the quantum rules no longer apply and the sender and receiver aren't made aware of the signal tampering.
Switzerland-based ID Quantique and Boston's Magiq Technologies, which made the quantum cryptography gear, welcomed the news as it will help them to shore up weaknesses in their encryption schemes. µ
Tags: Boffin watch
They found a mistake in the implementation of the system and exploited it. The underlying quantum encryption system is still fine.
It's roughly the same as attacking a swipe card system by hooking a 40,000 volt line up to the card reader. A naive implementation might fail "open", which is basically what happened to the hardware in this case.