tracking each individual bill as it is
used. They achieved this by requiring
that the owner of a bill modify the bill
before allowing the bank to verify it.
The modification is done in a special
way so that valid bills remain valid
but are otherwise randomized so that
the bank cannot tell them apart. This
scheme has the significant disadvantage that upon discovering a single
counterfeit bill, the bank is required
to immediately invalidate every bill it
has ever issued. In our opinion this
scheme therefore has limited practical applicability.
The idea of public-key quantum
money gained traction in the years that
followed. Aaronson proved a “
complexity-theoretic no-cloning theorem,” 1
which showed that even with access
to a verifier, a counterfeiter with limited computational resources cannot
copy an arbitrary state. Mosca and
Stebila proposed18 the idea of a quantum coin as distinct from a quantum
bill—each quantum coin of a given
denomination would be identical.
Using the complexity-theoretic no-cloning theorem they argued it might
be possible to implement a quantum
coin protocol but they did not give a
concrete implementation. Aaronson1
proposed the first concrete scheme
for public-key quantum money; however, this scheme was shown to be
insecure in Lutomirski et al. 16 In the
latter paper, the authors suggested
the idea of collision-free quantum
money. Unlike quantum coins, each
collision-free quantum bill has a
serial number and nobody, not even
the mint, can produce two bills with
the same serial number. This feature
can be useful to prevent the mint
from printing more money than it
says it is printing. The mint posts
a list of all serial numbers of every
quantum bill ever produced, and we
can be sure the mint produced at most
one bill for each serial number on the
list. In a subsequent paper, Farhi et
al. proposed a concrete scheme they
believed was both collision free and
secure against counterfeiting. 11
Here, we tell you how some of these
proposals work.
Wiesner’s Quantum Money
Wiesner’s original quantum money
scheme26 works as follows. To produce
the resurgence
of interest in
quantum money
is centered
around the idea
of public-key
quantum money.
a quantum bill using n qubits, the
mint first chooses n one-qubit states
randomly drawn from the set {|Sz = 1ñ,
|Sz = −1ñ, |Sx = 1ñ, |Sx = −1ñ}. The mint
then assigns that state a classical
serial number. A piece of quantum
money consists of the n qubit state and
its serial number. The mint keeps a
list of all serial numbers issued as well
as a description of which state corresponds to which serial number. When
you pay for something with a quantum
bill, the merchant sends the quantum
state and its serial number back to the
mint for verification. The mint looks
up the serial number and retrieves
the description of the corresponding
quantum state. Then the mint verifies
the given state is the state that goes
with the attached serial number. This
kind of money cannot be forged by
someone outside the mint. Since a
would-be forger has no knowledge of
the basis that each qubit was prepared
in, the quantum no-cloning theorem
says he or she cannot reliably copy the
n qubit quantum state (Figure 2).
The main weakness in Wiesner’s
scheme is that the merchant must
communicate with the bank to verify
each transaction. So this scheme,
although theoretically inspiring and
provably secure, would not be much
more powerful than credit cards.
Wiesner’s scheme is a private-key
quantum money scheme because the
mint must keep a private secret—the
complete description of the state—to
use for verification.
Challenges in Designing
Public-key Quantum Money
The resurgence of interest in quantum
money is centered around the idea
of public-key quantum money. As we
have discussed, a public-key quantum
money scheme would have the following properties. 16
1. The mint can mint it. That is,
there is an efficient algorithm to produce the quantum money state.
2. Anyone can verify it without communicating with the mint. That is,
there is an efficient measurement anyone can perform that accepts money
produced by the mint with high probability and minimal damage.
3. No one (except possibly the
mint) can copy it. That is, no one other than the mint can efficiently pro-