PRIVACY by DESIGN (cont.)
by Andy Greenburg / 01.06.16
“It’s been more than 30 years since David Chaum launched the ideas that would serve as much of the groundwork for anonymity online. In doing so, he also helped spark the debate that’s endured ever since, over the anarchic freedoms that digital secrecy enables—the conflict between privacy advocates and governments known today as the “crypto wars.” Now Chaum has returned with his first online privacy invention in more than a decade. And with it, he wants to bring those crypto wars to an end. At the Real World Crypto conference at Stanford University today, Chaum plans to present for the first time a new encryption scheme he calls PrivaTegrity. Like other tools Chaum has spent his long career developing, PrivaTegrity is designed to allow fully secret, anonymous communications that no eavesdropper can crack, whether a hacker or an intelligence agency. But PrivaTegrity, which Chaum’s been developing as a side project for the last two years along with a team of academic partners at Purdue, Radboud University in the Netherlands, Birmingham University and other schools, is meant to be both more secure than existing online anonymity systems like Tor or I2P and also more efficient; he claims it will be fast enough to work as a smartphone app with no perceptible delay.
Chaum wouldn’t comment on whether the project, which has yet to be fully coded and tested, would be commercialized or run as a non-profit, but he says an alpha version for Android is in development that functions as an instant-messaging app. In future versions, Chaum and his collaborators plan to add features like larger file sharing for photos and video, the ability to follow Twitter-like feeds, and even financial transactions, all under the cover of strong anonymity with untraceable pseudonyms. “It’s a way to create a separate online reality,” says Chaum, “One in which all the various things we now know people like to do online can be done in a lightweight manner under a completely different and new and very attractive privacy and security model.”
That ambitious privacy toolset aside, Chaum is also building into PrivaTegrity another feature that’s sure to be far more controversial: a carefully controlled backdoor that allows anyone doing something “generally recognized as evil” to have their anonymity and privacy stripped altogether. Whoever controls that backdoor within PrivaTegrity would have the power to decide who counts as “evil”—too much power, Chaum recognizes, for any single company or government. So he’s given the task to a sort of council system. When PrivaTegrity’s setup is complete, nine server administrators in nine different countries would all need to cooperate to trace criminals within the network and decrypt their communications. The result, Chaum argues, is a new approach that “breaks the crypto wars,” satisfying both the law enforcement agencies who argue that encryption offers a haven for criminals, and also those who argue that it’s necessary to hobble mass spying. “If you want a way to solve this apparent logjam, here it is,” says Chaum. “We don’t have to give up on privacy. We don’t have to allow terrorists and drug dealers to use it. We can have a civil society electronically without the possibility of covert mass surveillance.”
Chaum’s quest for a shield against Internet surveillance began before most of the world was even aware of the Internet at all. His inventions include the first-ever cryptocurrency, a 1990s venture known as DigiCash, and DC Nets, a scheme he invented in the early ’80s to allow theoretically perfect anonymity within a group of computers. But perhaps the most influential of Chaum’s privacy ideas was an earlier, simpler scheme he called a “mix network,” a term he coined in 1979.
Mix networks anonymize messages by encrypting them in layers and routing them through a series of computers that serve as intermediaries. Each of those middlemen machines collects messages in batches, shuffles them, strips off one layer of their encryption that only that computer can decrypt, and then passes them on to the next computer in the chain. The result is that no one, not even the individual intermediary computers themselves, can trace the messages from origin to destination. Today, anonymity tools inspired by mix networks are used by everyone from the nearly 2 million inhabitants of the Tor anonymity network—whose messages are routed through a sort of mutated mix network of thousands of volunteer machines—to Bitcoin spenders hiding drug transactions on the Dark Web.
With PrivaTegrity, Chaum is introducing a new kind of mix network he calls cMix, designed to be far more efficient than the layered encryption scheme he created decades ago. In his cMix setup, a smartphone communicates with PrivaTegrity’s nine servers when the app is installed to establish a series of keys that it shares with each server. When the phone sends a message, it encrypts the message’s data by multiplying it by that series of unique keys. Then the message is passed around all nine servers, with each one dividing out its secret key and multiplying the data with a random number. On a second pass through the nine servers, the message is put into a batch with other messages, and each server shuffles the batch’s order using a randomized pattern only that server knows, then multiplies the messages with another random number. Finally, the process is reversed, and as the message passes through the servers one last time, all of those random numbers are divided out and replaced with keys unique to the message’s intended recipient, who can then decrypt and read it.
Chaum argues that PrivaTegrity’s setup is more secure than Tor, for instance, which passes messages through three volunteer computers which may or may not be trusted. Unlike PrivaTegrity, Tor also doesn’t deliver its messages in batches, a decision designed to allow fast Web browsing. But that tradeoff means a spy who watches both ends of Tor’s network of intermediary computers might be able to identify the same message going in one at one place and coming out at another, a problem PrivaTegrity batch system is designed to solve. PrivaTegrity’s protocol will be speedier than past attempts at implementing mix networks, Chaum claims. That supposed efficiency comes from the fact that the collections of random numbers it uses, both before and after the messages are shuffled, can be precomputed and passed between the servers during moments when the servers are idle, instead of being created in real-time and slowing down conversations. And because the entire cMix process is a series of simple multiplications and divisions, it’s far faster than the public key computations necessary in older mix networks, says Aggelos Kiayas, a computer science professor at the University of Connecticut who’s reviewed Chaum’s system. “It is well known that mix nets can be better than Tor in terms of privacy…The real question is latency,” Kiayas writes in an email, cautioning that he can’t fully judge the scheme’s efficiency without seeing the final app. “PrivaTegrity appears to be a decisive step forward in this direction.”
A Backdoor Security Council
On top of those security and efficiency tricks, PrivaTegrity’s nine-server architecture—with a tenth that works as a kind of “manager” without access to any secret keys—also makes possible its unique backdoor decryption feature. No single server, or even eight of the nine servers working together, can trace or decrypt a message. But when all nine cooperate, they can combine their data to reconstruct a message’s entire path and divide out the random numbers they used to encrypt it. “It’s like a backdoor with nine different padlocks on it,” Chaum says.
The basic idea behind PrivaTegrity is that APT0 can’t *ever* compromise nine groups/computer systems. That is not a serious suggestion.
— Jacob Appelbaum (@ioerror) January 6, 2016
Chaum has yet to reveal the full list of the countries where PrivaTegrity would place its servers. But he suggests they’ll be in the jurisdiction of democratic governments, and names Switzerland, Canada and Iceland as examples. “It’s like the UN,” says Chaum. “I don’t think a single jurisdiction should be able to covertly surveil the planet…In this system, there’s an agreement on the rules, and then we can enforce them.” The mere mention of a “backdoor”—no matter how many padlocks, checks, and balances restrict it—is enough to send shivers down the spines of most of the crypto community. But Chaum’s approach represents a bold attempt to end the stalemate between staunch privacy advocates and officials like FBI director James Comey, CIA deputy director Michael Morrell and British Prime Minister David Cameron who have all opposed tech companies’ use of strong, end-to-end encryption. Comey, Cameron, and Morell have lashed out at firms like Apple and Whatsapp, for instance, for using systems in which even the company itself doesn’t possess the key to decrypt communications or stored data, and thus can’t cooperate with law enforcement. (Those same privacy features have earned the companies praise from privacy groups.) The debate between encryption fans and surveillance hawks has only intensified in the wake of ISIS’s attacks in Paris, and in last month’s Democratic presidential debate Hillary Clinton called for a “Manhattan-like Project” to develop a system that “would bring the government and the tech communities together.”2
Most encryption experts insist, however, that any backdoor would lead to abuse by hackers, if not by the very law enforcement or national security agencies it was created for. Chaum counters that spreading the keys to decrypt communications among nine servers would solve both of those problems, preventing abusive government surveillance and making his backdoor far harder to hack. He suggests that the servers’ administrator will eventually develop their own security protections and even distinct code to implement PrivaTegrity’s protocol, avoiding any single bug that could be common to all nine nodes. “These systems would be far more hardened than even corporate systems, and to abuse the backdoor you’d have to break all of them,” he says. Whether PrivaTegrity lives up to its efficiency and security promises will only become clear when the finished app is released, and Chaum himself, despite spending two years perfecting its crypto system, hasn’t even tried the final demo of the app’s private alpha. He remains cagey about naming a date for releasing the public beta and publishing its code so that it can be scoured for flaws, but he says there’s “no technical reason why it couldn’t be ready for the first quarter of 2016.” If PrivaTegrity’s reality matches Chaum’s descriptions of its potential, he hopes it could serve as a model for how other encryption systems can protect innocent people from spying without offering impunity to criminals. “You have to perfect the traceability of the evil people and the untraceability of the honest people,” says Chaum. “That’s how you break the apparent tradeoff, this standoff called the encryption wars.” For more technical information on the cMix idea that PrivaTegrity will use, here’s Chaum’s and his co-authors’ still-unpublished paper on the system:1
1Updated 1/6/2016 9:30am EST to embed the technical paper describing Privategrity’s cMix system.
2Correction 1/6/2016 3:30pm EST: An earlier version of the story misspelled the name of FBI director James Comey.
Anonymous Communications Network Called PrivaTegrity launched
by Vijay Prabhu / January 10, 2016
“His paper, called “cMix: Anonymization by High-Performance Scalable Mixing,” presents an evolution of the Mix Network concept, called cMix. The paper addresses several issues that governments and regular users had with the Mix Network. According to Chaum, now researchers plan to use this new cryptography protocol to build their own, more secure PrivaTegrity network, as an alternative to Tor. According to their research paper (page 1 – Section I, and page 4 – Chapter III-D), for each communications path established in a cMix network, the message sender creates connections with a series of trusted servers, with which it shares a series of keys. When the sender sends out a message, its data is multiplied with all the keys. As the message passes through each server, it is divided with each server’s corresponding key, but also multiplied again by a random number. Messages are then stored randomized in each server’s buffer. When the data needs to be retrieved and sent to the receiver, each server will retrieve the message from its random position, divide out the random numbers, and then multiply it with the recipient’s keys. When the data arrives on the recipient’s computer, their keys are used to divide the data and decrypt the message.
Chaum says that by moving most of the computational operations to the server, instead of the client, cMix achieves the same transfer speeds as Tor, but unlike its predecessor, it is not vulnerable to a series of “tagging” attacks. Tagging attacks rely on compromising Tor nodes, which on their own allow attackers to tag input slots with their output location. By using cMix’s setup, the protocol is not vulnerable to these type of attacks, unless the malicious actor compromises all the elements of the protocol, which is very rare. Additionally, researchers said that, in PrivaTegrity, tagging attacks are also blocked by how users set up keys with network nodes. “PrivaTegrity aims to provide privacy at a technical level that is not penetrable by nation states,” the researchers claim. “PrivaTegrity implements a new approach to user identification requiring each user to provide a small but different type of identifying information to each mix node.”
VULNERABILITY BAKED IN
Pioneer In Internet Anonymity Hands FBI A Huge Gift In Building Dangerous Backdoored Encryption System
by Mike Masnick / Jan 6th 2016
“I first came across cryptography pioneer David Chaum about a decade ago, during the debates about online voting. Many in the technology world were insisting that such things were impossible to do safely, but Chaum insisted he had come up with a way to do online voting safely (he’d also tried to do electronic money, DigiCash… unsuccessfully). Many people disagreed with Chaum and it led to some fairly epic discussions. It appears that Chaum is again making moves that are making many of his colleagues angry: specifically creating a backdoored encryption system.
Few doubt Chaum’s cryptography skills or pedigree. He was instrumental in the early days of computer cryptography and what anonymity we have online today owes a lot to Chaum. But his latest plan is… troublingUnfortunately, Chaum is both totally missing the point and playing right into the FBI’s hands. The argument of basically every other cryptographer is that building any encryption system is incredibly difficult — and introducing any sort of backdoor opens up massive and dangerous vulnerabilities — whether the original creators recognize it or not. The second you introduce a backdoor — even using Chaum’s weird “nine people in nine countries” system — you have introduced a vulnerability. A vulnerability that can and will be abused by others. You are introducing a security flaw. And that’s a massive security problem. Chaum’s bragging about this system totally misses this point: “If you want a way to solve this apparent logjam, here it is,” says Chaum. “We don’t have to give up on privacy. We don’t have to allow terrorists and drug dealers to use it. We can have a civil society electronically without the possibility of covert mass surveillance.”
That assumes that his system can’t be hacked. That’s a dangerous claim. Yes, the “key” is split into 9 pieces, but it’s still introducing a vulnerability and undermining the integrity of the system. And, worst of all, as ACLU security expert Chris Soghoian points out, this is little more than a huge political gift to the FBI, who can go back to their stupid claims that if technologists just work harder they can come up with a “solution” to the false problem of “going dark.” Similarly, you have politicians like Hillary Clinton insisting that if only techies come together with government they can “solve” the encryption/”going dark” issue. And now you can bet, without a doubt, that law enforcement and clueless politicians will start pointing to Chaum’s offering as an example of a “solution.”
Security experts: Backdoors weaken security. They’re a bad idea. Chaum: I’ve built a new system with a backdoor. FBI: See? It is possible.
— Christopher Soghoian (@csoghoian) January 6, 2016
But, as Soghoian points out, that misses the point. Chaum is creating a technology that is, by default, less secure and comes with vulnerabilities built in. It’s no secret that it’s possible to build backdoored encryption. Hell, just about anyone could do that. The “impossible” part that people are warning about is building such a system that is actually secure. Chaum’s is not. By default, it has vulnerabilities built in, and they will get exploited. And, even before the technology is exploited, the existence of this will be exploited by politicians and law enforcement to undermine arguments for strong encryption. And, of course, none of PrivaTegrity’s security claims have been checked or audited publicly at this point. Chaum admits that while the eventual plan will involve routing messages (multiple times) though nine servers in nine different countries, the prototype runs entirely on Amazon’s cloud computing infrastructure. Either way, at the very least, the system makes it clear that decrypting all such traffic requires attacking and compromising just nine servers. If you don’t think the NSA can do that, you haven’t been paying attention.”
PREVIOUSLY on #SPECTRE
RUBBER HOSE CRYPTO