Quantum Encryption: Empowering Code Breakers or Code Makers?

Dear George:

If, as you say, a quantum computer would be a “hacker’s dream,” I guess it would be a terrorist’s dream, too, right? If Osama Bin Laden could crack all codes that now protect digital data, he could scan Pentagon databases for targets while stealing funds from banks to pay for the attacks. Of course, the Pentagon and the banks would take countermeasures—and, meanwhile, the CIA might use quantum computers to break al-Qaida’s codes. The question is whether, when the smoke cleared, quantum computers would on balance have favored the bad guys or the good guys.

Certainly new technologies tend to empower some groups relative to others. The microcomputer revolution, including the Internet, seems to be basically pro-pluralism and anti-authoritarianism. But saying that now is easier than predicting it 40 years ago. Is it possible to predict, now, whether quantum technology will shift power toward the Bin Ladens of the world or to the Rumsfelds and Ashcrofts of the world? (I’ll leave aside the excellent question of how much technological power we’d really want the Ashcrofts to have.)

As I understand it, pre-quantum encryption technology is basically egalitarian. Thanks to such tools as public-key encryption—which you explain very clearly in A Shortcut Through Time—a person of meager means can easily send messages that are forbiddingly hard for even well-heeled code breakers to crack. In fact, the Clinton administration so feared the ability of criminals and rag-tag terrorists to harness this power that it hatched the ill-fated “Clipper Chip” plan, which would have given microchips government-issue codes, which the government, with a court order, could unlock to eavesdrop on phone calls or e-mails.

As you note, quantum computers, with their immense factoring power, would make things like public-key encryption suddenly vulnerable. In principle, I’d expect this development to favor the powerful over the less powerful, and the Ashcrofts over the terrorists. After all, powerful quantum computers would presumably be hugely expensive at first—and would have so few manufacturers, churning out so few copies, that their dissemination could be regulated. In fact, they might for a while emerge only from Pentagon-funded ventures.

So for a time we might have something like the Clipper-Chip scenario—an American government that could crack everyone’s codes. Then as quantum computing trickled down to the masses, things would even out: Everyone would have the power to break everyone’s codes.

Such fleeting concentrations of power are an old story with information technologies. The invention of writing in ancient states empowered ruling elites (who trained and employed the scribes); but the evolution of a user-friendly phonetic script and the attendant spread of literacy eroded this power. Mainframe computers gave power to governments and big corporations; then microcomputers gave power to small interest groups and upstart entrepreneurs.

However, I can imagine reasons this pattern wouldn’t repeat itself in the case of quantum code-breaking. As we’ve discussed, quantum technology can do more than break codes; it can also help protect codes by allowing you to send a decoding key from one place to another with no chance of its being intercepted. Or, in a different but roughly equivalent approach, quantum technology can let you send a key that will tell the recipient if anyone has managed to peek at it in transit—the approach taken by that already-on-the-market quantum-key system you mentioned in your last post.

So here’s one big question: When you are assured that your key is secure, can you then use a code that isn’t vulnerable to the peculiar code-cracking powers of a quantum computer—e.g., a code whose cracking didn’t involve factoring large numbers? For all I know, a quantum computer, faced with any kind of code, could infer the key from the encrypted message via some kind of superfast trial and error. (Background for people who haven’t read your book: With the currently popular public-key encryption, you don’t have to send the key securely, because possession of the key alone doesn’t help an eavesdropper crack the code—unless said eavesdropper can solve massive factoring problems.)

Here’s another big question: Even if the answer to that last question is no, could quantum computers in some other way wind up empowering code makers as much as code breakers?

If the answer to either question is yes, then I can imagine us winding up with the status quo ante, as quantum technologies fended off the code-breaking threat posed by future quantum computers. But if the answer to both questions is no, then technological history might repeat its familiar pattern: The Ashcrofts of the world will have their days of dominance, but then the dominant technology will spread beyond their control. So we’ll have moved from the current kind of egalitarianism, in which everyone can in principle send secure messages, to a new kind of egalitarianism, in which everyone can in principle break everyone else’s codes.

George, I’m confident that I don’t know enough to have gotten through the previous seven paragraphs without making significant errors of fact or logic. (One haunting afterthought: I may be glossing over a big difference between the challenge of sending a secure message and the challenge of keeping a stationary database secure.) America is counting on you to straighten me out and clarify the future, so that we can decide how much of the budget deficit to devote to a crash program to build quantum computers.

But first let me briefly revert to an earlier topic and give my final defense of my (and Einstein’s, dammit!) insistence that “spooky action at a distance” is an apt description of quantum entanglement: When I perform an action in New York that leads to the creation of a new piece of information, and that information instantaneously shows up in New Mexico as well as New York, that’s a) action; b) at a distance; and c) spooky.

But not as spooky as al-Qaida being able to crack all our codes. So do enlighten us on that last point, please. And thanks again for a great book.