Entries Tagged "backdoors"

Page 17 of 22

iPhone Encryption and the Return of the Crypto Wars

Last week, Apple announced that it is closing a serious security vulnerability in the iPhone. It used to be that the phone’s encryption only protected a small amount of the data, and Apple had the ability to bypass security on the rest of it.

From now on, all the phone’s data is protected. It can no longer be accessed by criminals, governments, or rogue employees. Access to it can no longer be demanded by totalitarian governments. A user’s iPhone data is now more secure.

To hear US law enforcement respond, you’d think Apple’s move heralded an unstoppable crime wave. See, the FBI had been using that vulnerability to get into people’s iPhones. In the words of cyberlaw professor Orin Kerr, “How is the public interest served by a policy that only thwarts lawful search warrants?”

Ah, but that’s the thing: You can’t build a backdoor that only the good guys can walk through. Encryption protects against cybercriminals, industrial competitors, the Chinese secret police and the FBI. You’re either vulnerable to eavesdropping by any of them, or you’re secure from eavesdropping from all of them.

Backdoor access built for the good guys is routinely used by the bad guys. In 2005, some unknown group surreptitiously used the lawful-intercept capabilities built into the Greek cell phone system. The same thing happened in Italy in 2006.

In 2010, Chinese hackers subverted an intercept system Google had put into Gmail to comply with US government surveillance requests. Back doors in our cell phone system are currently being exploited by the FBI and unknown others.

This doesn’t stop the FBI and Justice Department from pumping up the fear. Attorney General Eric Holder threatened us with kidnappers and sexual predators.

The former head of the FBI’s criminal investigative division went even further, conjuring up kidnappers who are also sexual predators. And, of course, terrorists.

FBI Director James Comey claimed that Apple’s move allows people to “place themselves beyond the law” and also invoked that now overworked “child kidnapper.” John J. Escalante, chief of detectives for the Chicago police department now holds the title of most hysterical: “Apple will become the phone of choice for the pedophile.”

It’s all bluster. Of the 3,576 major offenses for which warrants were granted for communications interception in 2013, exactly one involved kidnapping. And, more importantly, there’s no evidence that encryption hampers criminal investigations in any serious way. In 2013, encryption foiled the police nine times, up from four in 2012­—and the investigations proceeded in some other way.

This is why the FBI’s scare stories tend to wither after public scrutiny. A former FBI assistant director wrote about a kidnapped man who would never have been found without the ability of the FBI to decrypt an iPhone, only to retract the point hours later because it wasn’t true.

We’ve seen this game before. During the crypto wars of the 1990s, FBI Director Louis Freeh and others would repeatedly use the example of mobster John Gotti to illustrate why the ability to tap telephones was so vital. But the Gotti evidence was collected using a room bug, not a telephone tap. And those same scary criminal tropes were trotted out then, too. Back then we called them the Four Horsemen of the Infocalypse: pedophiles, kidnappers, drug dealers, and terrorists. Nothing has changed.

Strong encryption has been around for years. Both Apple’s FileVault and Microsoft’s BitLocker encrypt the data on computer hard drives. PGP encrypts e-mail. Off-the-Record encrypts chat sessions. HTTPS Everywhere encrypts your browsing. Android phones already come with encryption built-in. There are literally thousands of encryption products without back doors for sale, and some have been around for decades. Even if the US bans the stuff, foreign companies will corner the market because many of us have legitimate needs for security.

Law enforcement has been complaining about “going dark” for decades now. In the 1990s, they convinced Congress to pass a law requiring phone companies to ensure that phone calls would remain tappable even as they became digital. They tried and failed to ban strong encryption and mandate back doors for their use. The FBI tried and failed again to ban strong encryption in 2010. Now, in the post-Snowden era, they’re about to try again.

We need to fight this. Strong encryption protects us from a panoply of threats. It protects us from hackers and criminals. It protects our businesses from competitors and foreign spies. It protects people in totalitarian governments from arrest and detention. This isn’t just me talking: The FBI also recommends you encrypt your data for security.

As for law enforcement? The recent decades have given them an unprecedented ability to put us under surveillance and access our data. Our cell phones provide them with a detailed history of our movements. Our call records, e-mail history, buddy lists, and Facebook pages tell them who we associate with. The hundreds of companies that track us on the Internet tell them what we’re thinking about. Ubiquitous cameras capture our faces everywhere. And most of us back up our iPhone data on iCloud, which the FBI can still get a warrant for. It truly is the golden age of surveillance.

After considering the issue, Orin Kerr rethought his position, looking at this in terms of a technological-legal trade-off. I think he’s right.

Given everything that has made it easier for governments and others to intrude on our private lives, we need both technological security and legal restrictions to restore the traditional balance between government access and our security/privacy. More companies should follow Apple’s lead and make encryption the easy-to-use default. And let’s wait for some actual evidence of harm before we acquiesce to police demands for reduced security.

This essay previously appeared on CNN.com

EDITED TO ADD (10/6): Three more essays worth reading. As is this on all the other ways Apple and the government have to get at your iPhone data.

And an Washington Post editorial manages to say this:

How to resolve this? A police “back door” for all smartphones is undesirable—a back door can and will be exploited by bad guys, too. However, with all their wizardry, perhaps Apple and Google could invent a kind of secure golden key they would retain and use only when a court has approved a search warrant.

Because a “secure golden key” is completely different from a “back door.”

EDITED TO ADD (10/7): Another essay.

EDITED TO ADD (10/9): Three more essays that are worth reading.

EDITED TO ADD (10/12): Another essay.

Posted on October 6, 2014 at 6:50 AM • View Comments

Could Keith Alexander's Advice Possibly Be Worth $600K a Month?

Ex-NSA director Keith Alexander has his own consulting company: IronNet Cybersecurity Inc. His advice does not come cheap:

Alexander offered to provide advice to Sifma for $1 million a month, according to two people briefed on the talks. The asking price later dropped to $600,000, the people said, speaking on condition of anonymity because the negotiation was private.

Alexander declined to comment on the details, except to say that his firm will have contracts “in the near future.”

Kenneth Bentsen, Sifma’s president, said at a Bloomberg Government event yesterday in Washington that “cybersecurity is probably our number one priority” now that most regulatory changes imposed after the 2008 credit crisis have been absorbed.

SIFMA is the Securities Industry and Financial Markets Association. Think of how much actual security they could buy with that $600K a month. Unless he’s giving them classified information.

Digby:

But don’t worry, everything Alexander knows will only benefit the average American like you and me. There’s no reason to suspect that he is trading his high level of inside knowledge to benefit a bunch of rich people all around the globe. Because patriotism.

Or, as Recode.net said: “For another million, I’ll show you the back door we put in your router.”

EDITED TO ADD (7/13): Rep. Alan Grayson is suspicious.

Posted on June 24, 2014 at 2:30 PM • View Comments

Internet Subversion

In addition to turning the Internet into a worldwide surveillance platform, the NSA has surreptitiously weakened the products, protocols, and standards we all use to protect ourselves. By doing so, it has destroyed the trust that underlies the Internet. We need that trust back.

Trust is inherently social. It is personal, relative, situational, and fluid. It is not uniquely human, but it is the underpinning of everything we have accomplished as a species. We trust other people, but we also trust organizations and processes. The psychology is complex, but when we trust a technology, we basically believe that it will work as intended.

This is how we technologists trusted the security of the Internet. We didn’t have any illusions that the Internet was secure, or that governments, criminals, hackers, and others couldn’t break into systems and networks if they were sufficiently skilled and motivated. We didn’t trust that the programmers were perfect, that the code was bug-free, or even that our crypto math was unbreakable. We knew that Internet security was an arms race, and the attackers had most of the advantages.

What we trusted was that the technologies would stand or fall on their own merits.

We now know that trust was misplaced. Through cooperation, bribery, threats, and compulsion, the NSA—and the United Kingdom’s GCHQ—forced companies to weaken the security of their products and services, then lie about it to their customers.

We know of a few examples of this weakening. The NSA convinced Microsoft to make some unknown changes to Skype in order to make eavesdropping on conversations easier. The NSA also inserted a degraded random number generator into a common standard, then worked to get that generator used more widely.

I have heard engineers working for the NSA, FBI, and other government agencies delicately talk around the topic of inserting a “backdoor” into security products to allow for government access. One of them told me, “It’s like going on a date. Sex is never explicitly mentioned, but you know it’s on the table.” The NSA’s SIGINT Enabling Project has a $250 million annual budget; presumably it has more to show for itself than the fragments that have become public. Reed Hundt calls for the government to support a secure Internet, but given its history of installing backdoors, why would we trust claims that it has turned the page?

We also have to assume that other countries have been doing the same things. We have long believed that networking products from the Chinese company Huawei have been backdoored by the Chinese government. Do we trust hardware and software from Russia? France? Israel? Anywhere?

This mistrust is poison. Because we don’t know, we can’t trust any of them. Internet governance was largely left to the benign dictatorship of the United States because everyone more or less believed that we were working for the security of the Internet instead of against it. But now that system is in turmoil. Foreign companies are fleeing US suppliers because they don’t trust American firms’ security claims. Far worse governments are using these revelations to push for a more isolationist Internet, giving them more control over what their citizens see and say.

All so we could eavesdrop better.

There is a term in the NSA: “nobus,” short for “nobody but us.” The NSA believes it can subvert security in such a way that only it can take advantage of that subversion. But that is hubris. There is no way to determine if or when someone else will discover a vulnerability. These subverted systems become part of our infrastructure; the harms to everyone, once the flaws are discovered, far outweigh the benefits to the NSA while they are secret.

We can’t both weaken the enemy’s networks and protect our own. Because we all use the same products, technologies, protocols, and standards, we either allow everyone to spy on everyone, or prevent anyone from spying on anyone. By weakening security, we are weakening it against all attackers. By inserting vulnerabilities, we are making everyone vulnerable. The same vulnerabilities used by intelligence agencies to spy on each other are used by criminals to steal your passwords. It is surveillance versus security, and we all rise and fall together.

Security needs to win. The Internet is too important to the world—and trust is too important to the Internet—to squander it like this. We’ll never get every power in the world to agree not to subvert the parts of the Internet they control, but we can stop subverting the parts we control. Most of the high-tech companies that make the Internet work are US companies, so our influence is disproportionate. And once we stop subverting, we can credibly devote our resources to detecting and preventing subversion by others.

This essay previously appeared in the Boston Review.

Posted on May 12, 2014 at 6:26 AM • View Comments

NSA Hacks Huawei

Both Der Spiegel and the New York Times are reporting that the NSA has hacked Huawei pretty extensively, getting copies of the company’s products’ source code and most of the e-mail from the company. Aside from being a pretty interesting story about the operational capabilities of the NSA, it exposes some pretty blatant US government hypocrisy on this issue. As former Bush administration official (and a friend of mine) Jack Goldsmith writes:

The Huawei revelations are devastating rebuttals to hypocritical U.S. complaints about Chinese penetration of U.S. networks, and also make USG protestations about not stealing intellectual property to help U.S. firms’ competitiveness seem like the self-serving hairsplitting that it is. (I have elaborated on these points many times and will not repeat them here.) “The irony is that exactly what they are doing to us is what they have always charged that the Chinese are doing through us,” says a Huawei Executive.

This isn’t to say that the Chinese are not targeting foreign networks through Huawei equipment; they almost certainly are.

Posted on March 24, 2014 at 12:51 PM • View Comments

Computer Network Exploitation vs. Computer Network Attack

Back when we first started getting reports of the Chinese breaking into U.S. computer networks for espionage purposes, we described it in some very strong language. We called the Chinese actions cyber–attacks. We sometimes even invoked the word cyberwar, and declared that a cyber-attack was an act of war.

When Edward Snowden revealed that the NSA has been doing exactly the same thing as the Chinese to computer networks around the world, we used much more moderate language to describe U.S. actions: words like espionage, or intelligence gathering, or spying. We stressed that it’s a peacetime activity, and that everyone does it.

The reality is somewhere in the middle, and the problem is that our intuitions are based on history.

Electronic espionage is different today than it was in the pre-Internet days of the Cold War. Eavesdropping isn’t passive anymore. It’s not the electronic equivalent of sitting close to someone and overhearing a conversation. It’s not passively monitoring a communications circuit. It’s more likely to involve actively breaking into an adversary’s computer network—be it Chinese, Brazilian, or Belgian—and installing malicious software designed to take over that network.

In other words, it’s hacking. Cyber-espionage is a form of cyber-attack. It’s an offensive action. It violates the sovereignty of another country, and we’re doing it with far too little consideration of its diplomatic and geopolitical costs.

The abbreviation-happy U.S. military has two related terms for what it does in cyberspace. CNE stands for “computer network exploitation.” That’s spying. CNA stands for “computer network attack.” That includes actions designed to destroy or otherwise incapacitate enemy networks. That’s—among other things—sabotage.

CNE and CNA are not solely in the purview of the U.S.; everyone does it. We know that other countries are building their offensive cyberwar capabilities. We have discovered sophisticated surveillance networks from other countries with names like GhostNet, Red October, The Mask. We don’t know who was behind them—these networks are very difficult to trace back to their source—but we suspect China, Russia, and Spain, respectively. We recently learned of a hacking tool called RCS that’s used by 21 governments: Azerbaijan, Colombia, Egypt, Ethiopia, Hungary, Italy, Kazakhstan, Korea, Malaysia, Mexico, Morocco, Nigeria, Oman, Panama, Poland, Saudi Arabia, Sudan, Thailand, Turkey, UAE, and Uzbekistan.

When the Chinese company Huawei tried to sell networking equipment to the U.S., the government considered that equipment a “national security threat,” rightly fearing that those switches were backdoored to allow the Chinese government both to eavesdrop and attack US networks. Now we know that the NSA is doing the exact same thing to American–made equipment sold in China, as well as to those very same Huawei switches.

The problem is that, from the point of view of the object of an attack, CNE and CNA look the same as each other, except for the end result. Today’s surveillance systems involve breaking into the computers and installing malware, just as cybercriminals do when they want your money. And just like Stuxnet: the U.S./Israeli cyberweapon that disabled the Natanz nuclear facility in Iran in 2010.

This is what Microsoft’s General Counsel Brad Smith meant when he said: “Indeed, government snooping potentially now constitutes an ‘advanced persistent threat,’ alongside sophisticated malware and cyber attacks.”

When the Chinese penetrate U.S. computer networks, which they do with alarming regularity, we don’t really know what they’re doing. Are they modifying our hardware and software to just eavesdrop, or are they leaving :logic bombs” that could be triggered to do real damage at some future time? It can be impossible to tell. As a 2011 EU cybersecurity policy document stated (page 7):

…technically speaking, CNA requires CNE to be effective. In other words, what may be preparations for cyberwarfare can well be cyberespionage initially or simply be disguised as such.

We can’t tell the intentions of the Chinese, and they can’t tell ours, either.

Much of the current debate in the U.S. is over what the NSA should be allowed to do, and whether limiting the NSA somehow empowers other governments. That’s the wrong debate. We don’t get to choose between a world where the NSA spies and one where the Chinese spy. Our choice is between a world where our information infrastructure is vulnerable to all attackers or secure for all users.

As long as cyber-espionage equals cyber-attack, we would be much safer if we focused the NSA’s efforts on securing the Internet from these attacks. True, we wouldn’t get the same level of access to information flows around the world. But we would be protecting the world’s information flows—including our own—from both eavesdropping and more damaging attacks. We would be protecting our information flows from governments, nonstate actors, and criminals. We would be making the world safer.

Offensive military operations in cyberspace, be they CNE or CNA, should be the purview of the military. In the U.S., that’s CyberCommand. Such operations should be recognized as offensive military actions, and should be approved at the highest levels of the executive branch, and be subject to the same international law standards that govern acts of war in the offline world.

If we’re going to attack another country’s electronic infrastructure, we should treat it like any other attack on a foreign country. It’s no longer just espionage, it’s a cyber-attack.

This essay previously appeared on TheAtlantic.com.

Posted on March 10, 2014 at 6:46 AM • View Comments

Was the iOS SSL Flaw Deliberate?

Last October, I speculated on the best ways to go about designing and implementing a software backdoor. I suggested three characteristics of a good backdoor: low chance of discovery, high deniability if discovered, and minimal conspiracy to implement.

The critical iOS vulnerability that Apple patched last week is an excellent example. Look at the code. What caused the vulnerability is a single line of code: a second “goto fail;” statement. Since that statement isn’t a conditional, it causes the whole procedure to terminate.

The flaw is subtle, and hard to spot while scanning the code. It’s easy to imagine how this could have happened by error. And it would have been trivially easy for one person to add the vulnerability.

Was this done on purpose? I have no idea. But if I wanted to do something like this on purpose, this is exactly how I would do it.

EDITED TO ADD (2/27): If the Apple auditing system is any good, they would be able to trace this errant goto line not just to the source-code check-in details, but to the specific login that made the change. And they would quickly know whether this was just an error, or a deliberate change by a bad actor. Does anyone know what’s going on inside Apple?

EDITED TO ADD (2/27): Steve Bellovin has a pair of posts where he concludes that if this bug is enemy action, it’s fairly clumsy and unlikely to be the work of professionals.

Posted on February 27, 2014 at 6:03 AM • View Comments

STUCCOMONTANA: NSA Exploit of the Day

Today’s implant from the NSA’s Tailored Access Operations (TAO) group implant catalog:

STUCCOMONTANA

(TS//SI//REL) STUCCOMONTANA provides persistence for DNT implants. The DNT implant will survive an upgrade or replacement of the operating system—including physically replacing the router’s compact flash card.

(TS//SI//REL) Currently, the intended DNT Implant to persist is VALIDATOR, which must be run as a user process on the target operating system. The vector of attack is the modification of the target’s BIOS. The modification will add the necessary software to the BIOS and modify its software to execute the SIERRAMONTANA implant at the end of its native System Management Mode (SMM) handler.

(TS//SI//REL) STUCCOMONTANA must support all modern versions of JUNOS, which is a version of FreeBSD customized by Juniper. Upon system boot, the JUNOS operating system is modified in memory to run the implant, and provide persistent kernel modifications to support implant execution.

(TS//SI//REL) STUCCOMONTANA is the cover term for the persistence technique to deploy a DNT implant to Juniper T-Series routers.

Unit Cost: $

Status: (U//FOUO) STUCCOMONTANA under development and is expected to be released by 30 November 2008.

Page, with graphics, is here. General information about TAO and the catalog is here.

In the comments, feel free to discuss how the exploit works, how we might detect it, how it has probably been improved since the catalog entry in 2008, and so on.

Posted on January 17, 2014 at 2:06 PM • View Comments

SIERRAMONTANA: NSA Exploit of the Day

Today’s implant from the NSA’s Tailored Access Operations (TAO) group implant catalog:

SIERRAMONTANA

(TS//SI//REL) SIERRAMONTANA provides persistence for DNT implants. The DNT implant will survive an upgrade or replacement of the operating system—including physically replacing the router’s compact flash card.

(TS//SI//REL) Currently, the intended DNT Implant to persist is VALIDATOR, which must be run as a user process on the target operating system. The vector of attack is the modification of the target’s BIOS. The modification will add the necessary software to the BIOS and modify its software to execute the SIERRAMONTANA implant at the end of its native System Management Mode (SMM) handler.

(TS//SI//REL) SIERRAMONTANA must support all modern versions of JUNOS, which is a version of FreeBSD customized by Juniper. Upon system boot, the JUNOS operating system is modified in memory to run the implant, and provide persistent kernel modifications to support implant execution.

(TS//SI//REL) SIERRAMONTANA is the cover term for the persistence technique to deploy a DNT implant to Juniper M-Series routers.

Unit Cost: $

Status: (U//FOUO) SIERRAMONTANA under development and is expected to be released by 30 November 2008.

Page, with graphics, is here. General information about TAO and the catalog is here.

We have already seen the codename VALIDATOR. It’s the code name for a default, or basic, NSA exploit. It’s the exploit that FOXACID defaults to using.

In the comments, feel free to discuss how the exploit works, how we might detect it, how it has probably been improved since the catalog entry in 2008, and so on.

Posted on January 16, 2014 at 2:00 PM • View Comments

SCHOOLMONTANA: NSA Exploit of the Day

Today’s implant from the NSA’s Tailored Access Operations (TAO) group implant catalog:

SCHOOLMONTANA

(TS//SI//REL) SCHOOLMONTANA provides persistence for DNT implants. The DNT implant will survive an upgrade or replacement of the operating system—including physically replacing the router’s compact flash card.

(TS//SI//REL) Currently, the intended DNT Implant to persist is VALIDATOR, which must be run as a user process on the target operating system. The vector of attack is the modification of the target’s BIOS. The modification will add the necessary software to the BIOS and modify its software to execute the SCHOOLMONTANA implant at the end of its native System Management Mode (SMM) handler.

(TS//SI//REL) SCHOOLMONTANA must support all modern versions of JUNOS, which is a version of FreeBSD customized by Juniper. Upon system boot, the JUNOS operating system is modified in memory to run the implant, and provide persistent kernel modifications to support implant execution.

(TS//SI//REL) SCHOOLMONTANA is the cover term for the persistence technique to deploy a DNT implant to Juniper J-Series routers.

Status: (U//FOUO) SCHOOLMONTANA completed and released by ANT May 30, 2008. It is ready for deployment.

Page, with graphics, is here. General information about TAO and the catalog is here.

In the comments, feel free to discuss how the exploit works, how we might detect it, how it has probably been improved since the catalog entry in 2008, and so on.

Posted on January 15, 2014 at 2:56 PM • View Comments

1 … 15 16 17 18 19 … 22

Sidebar photo of Bruce Schneier by Joe MacInnis.