IBC
article here
While Google forges ahead with unlocking the potential of
its Willow quantum computing chip, cybersecurity experts warn that further
breakthroughs in the field could catch a digital ecosystem built on crypto
security off guard.
Imagine a world where, instead of relying on your own
carelessness with house keys to gain entry, burglars only need to examine the
lock to cut their own keys. Scale that to an enterprise level, and this is what
is believed will happen to every digitally encrypted system on the planet.
“Existing cryptographic systems could be broken by quantum
almost overnight,” warns Julian Van Velzen, CTIO and Head of the Quantum Lab at
Cap Gemini.
Dubbed the ‘quantum apocalypse’, it could happen as soon as
2030.
“Quantum completely changes the way we need to think about
security,” concurs Bob Oates, Associate Director and cybersecurity specialist
at Cambridge Consultants.
Robin Boldon, Head of Product at anti-piracy solutions
provider Friend MTS, equally urges caution. “Encryption has launched new
business opportunities for our industry but has also exposed some of its
weaknesses by powering piracy at a scale never seen before. If quantum
technology advances in a negative way, you could see how it could unravel an
awful lot of businesses that are based on current cryptographic techniques.”
RSA and ECC
Encryption prevents the wrong information from getting into
the wrong hands, prevents private communications and data from being
manipulated, and gives us assurance that the company or person we are
transacting with is who we think they are. The problem is that prevailing
methods of cryptography – RSA and ECC – are out of date and
on the verge of being irrevocably cracked open by advances in quantum
computing.
“Existing crypto systems were invented in the ‘70s and ‘80s
and they are ageing,” says Mark Pecen, Chair of a new task force established
by the European Telecommunications Standards Institute (ETSI) to address
quantum security. “Combine that with the incredible evolution of quantum, and
the risk is real. We have to keep a step ahead of criminals because they will
attack from everywhere.”
RSA and ECC are among the most commonly used cryptographic
systems used to secure everything from military secrets to the emails we send
daily. Our PCs and mobile phones have specific hardware acceleration for this
cryptography built into them. Crucially, these are asymmetric systems that
require a public key and a private key to encrypt and decrypt data, and it was
thought that it would take millions of years for even the most powerful
supercomputer to crack the math.
Now it is theorised that a suitably capable quantum computer
could do it in hours.
It’s not so much that quantum computing can perform
calculations faster, but by enabling combinations of binary code (1 and 0) to
exist in multiple states (or superpositions) at the same time, it computes
problems in a fundamentally different way than serial or linear classical math.
Last year, Google announced that its latest
quantum chip had taken less than five minutes to solve a problem that
would have taken the world’s fastest conventional supercomputer ten
septillion years to complete. The breakthrough Google claimed, “brings us closer
to running practical, commercially-relevant algorithms that can’t be replicated
on conventional computers.” The company is now working with UK researchers to
come up with innovative use cases for its Willow quantum chip.
A matter of time
A quantum computer with the power to crack RSA does not yet
exist, but even the most sceptical of experts believe it to be a matter of
time. The consensus is that this will happen within the decade, meaning that
every company’s digital infrastructure is at risk if it is not upgraded to be
quantum safe.
“The date is contentious, but there are credible estimates
that between 2030 and the late 2040s RSA will fall,” says Oates. “Once that’s
fallen, the other [classical crypto systems] will be very close behind. It all
depends on the pace of quantum computing.”
“Let's say you have a computer centre with 100 CPUs. If you
want to double the power, you add another hundred CPUs,” says Pecen. “If you
have 100 qubits of quantum compute, and you want to double your power, you just
add 1 qubit. The capacity of this technology is exploding.”
Cybercriminals and state actors are already thought to be
harvesting data and storing it with a view to unlocking it with quantum crypto
down the line.
“Any kind of information that's passed cryptographically is
potentially at risk of being hacked,” warns Daryl Flack, Partner at Avella
Security. “Ultimately, companies need to start actively doing something now,
and the problem is that lots of people don't know about it. They are too caught
up in the AI wave. They're relying on big vendors to do the job for them, and
whilst that may be the case, you still need to understand your risk to
understand where your crypto is, what algorithms you're using, who you're
reliant upon, and ultimately, what service providers you might need to call on
to tackle this impending challenge.”
Post-quantum cryptography
The US government’s technical standards institute has
taken the lead. Last year it published three groups of algorithms that it
deemed to be resilient to quantum attack. It also mandated that US companies
outline a plan to migrate to post-quantum cryptography (PQC) using those
algorithms by 2027. Both the EU and the UK have almost identical legislative
clocks using the same algorithmic tools.
The EU advises member states to begin
transitioning to PQC by “no later than” 2026, that high-risk systems - meaning
utilities, telecom, finance and government infrastructure - must be secured by
2030 and that all other systems should have been upgraded by 2035.
“2035 may sound like a long way away but the lead times for
what will be required demand action now,” Flack says. “Some of this is going to
require new hardware. Device manufacturers in devices need to adjust their
manufacturing processes, build new devices, source new chips, test them, deploy
them, and if you've got estates with tens of thousands or millions of devices,
then replacing those by the beginning of 2030 will be challenging.”
Quantum and media
Amazon, Hewlett-Packard Enterprise, InterDigital and Pecen’s
own company EigenQ are among 25 members on the ETSI quantum task force.
“Governments understand the urgency, but I don’t think many
in broadcast and media do,” says Pecen. “CIOs at companies outside of media
that I have spoken with say that the first priority is to replace ageing
systems with quantum safe technology.”
“Cryptography is a specialism within a specialism,” says
Oates. “Your average IT security person might know some cryptography, but
they're not going to be able to really deep dive.”
The UK’s National Cyber Security Centre has certified a
number of companies to act as consultants for preparing to move to PQC. They
include Avella Security, Cambridge Consultants and Cap Gemini.
“Every company has supply chains with different vendors and
so it will be crucial to think about what kind of cryptography they use,”
advises Van Velzen. “Are they ready to become quantum safe? Do they have a
cryptographic roadmap? Can I put in appropriate contractual terms for all the
different vendors and products that I have? What are my peers doing and can we
agree on standards to sure that that the whole ecosystem is ready to migrate in
a sensible way?”
The quantum threat is on the radar of media security
specialist Synamedia, but then Alain Durand, its Senior Director, was a
cryptography engineer earlier in his career. “I think the initial risk will be
to break customer credit cards rather than media content but we always have to
look to the latest techniques and adapt,” he says. “Since cryptography is used
to secure content in a DRM or in conditional access, there will be a need for
an upgrade when quantum compute arrives.”
According to Oates, quantum computers enable broadly four
different types of attack: reading information they illegally possess, sending
malicious messages, faking digital signatures and being able to push out
software (malware) by pretending to be a supplier that you trust.
We know about advances in quantum computing from the work of
big tech giants like Google and Microsoft, but it’s the leaps forward we don’t
know about which should be concerning.
“Quantum computers will be available to break current
cryptographic systems sometime in the next decade but those machines won't be
cheap, and it won’t be on a laptop,” says Van Velzen. “It will be a
football-sized machine with millions of small parts and it will cost millions
of dollars to run an algorithm to break RSA in the beginning. Initially, it
would not be something that everyone has access to, but it would absolutely be
nation-states most likely to be able to perform such attacks.”
While nation-states are in a race to develop the technology
to spy on military communications, organised criminals are not far behind.
“We're already seeing some big tech players talking about renting quantum
compute in the cloud for a few hours at a time,” says Oates. “Once that becomes
commonplace and cost-effective, then there is a risk that organised criminals
will try and do the same.”
State of play
The first principles of quantum computing have been tested
and quantum processors are able to demonstrate interaction between qubits and
perform some small computations. According to Van Velzen, “We are now seeing
companies starting to scale.”
These include Infleqtion in Oxfordshire recipient
of a £2.2m fund from the UK government, and PsiQuantum in Silicon
Valley, which recently collected a $1bn investment to build what it claims will
be the world's first commercially useful, fault-tolerant quantum computer.
“The technological advance most people talk about is how
many qubits a quantum computer has, and while that is a measure, there are lots
of other things to consider,” says Oates. “The error rate is especially
important because the bigger the error rate, the more redundancy you need,
which means you need more qubits, which is a multiplying factor of how big the
quantum computer needs to be.” Other parameters such as connectivity are
equally valuable in connecting each qubit to every other qubit. “There is incredible
work going on into building bigger and better computers and every time there's
a massive breakthrough, it does shave a few years off the estimates for when it
becomes viable,” Oates says.
The media industry is urged to investigate the issue now, to
catalogue its cryptographic systems and prioritise an upgrade path. Inevitably,
this incurs an investment.
“Change will be required to the encryption algorithms, but
that won’t change the fundamental problem or the response to the challenge we
face today,” says Durand.
“Pirates will always find a way to steal.”
No comments:
Post a Comment