Entries Tagged "TLS"

Page 1 of 2

The NSA Warns of TLS Inspection

The NSA has released a security advisory warning of the dangers of TLS inspection:

Transport Layer Security Inspection (TLSI), also known as TLS break and inspect, is a security process that allows enterprises to decrypt traffic, inspect the decrypted content for threats, and then re-encrypt the traffic before it enters or leaves the network. Introducing this capability into an enterprise enhances visibility within boundary security products, but introduces new risks. These risks, while not inconsequential, do have mitigations.

[…]

The primary risk involved with TLSI’s embedded CA is the potential abuse of the CA to issue unauthorized certificates trusted by the TLS clients. Abuse of a trusted CA can allow an adversary to sign malicious code to bypass host IDS/IPSs or to deploy malicious services that impersonate legitimate enterprise services to the hosts.

[…]

A further risk of introducing TLSI is that an adversary can focus their exploitation efforts on a single device where potential traffic of interest is decrypted, rather than try to exploit each location where the data is stored.Setting a policy to enforce that traffic is decrypted and inspected only as authorized, and ensuring that decrypted traffic is contained in an out-of-band, isolated segment of the network prevents unauthorized access to the decrypted traffic.

[…]

To minimize the risks described above, breaking and inspecting TLS traffic should only be conducted once within the enterprise network. Redundant TLSI, wherein a client-server traffic flow is decrypted, inspected, and re-encrypted by one forward proxy and is then forwarded to a second forward proxy for more of the same,should not be performed.Inspecting multiple times can greatly complicate diagnosing network issues with TLS traffic. Also, multi-inspection further obscures certificates when trying to ascertain whether a server should be trusted. In this case, the “outermost” proxy makes the decisions on what server certificates or CAs should be trusted and is the only location where certificate pinning can be performed.Finally, a single TLSI implementation is sufficient for detecting encrypted traffic threats; additional TLSI will have access to the same traffic. If the first TLSI implementation detected a threat, killed the session, and dropped the traffic, then additional TLSI implementations would be rendered useless since they would not even receive the dropped traffic for further inspection. Redundant TLSI increases the risk surface, provides additional opportunities for adversaries to gain unauthorized access to decrypted traffic, and offers no additional benefits.

Nothing surprising or novel. No operational information about who might be implementing these attacks. No classified information revealed.

News article.

Posted on November 22, 2019 at 6:16 AMView Comments

New Reductor Nation-State Malware Compromises TLS

Kaspersky has a detailed blog post about a new piece of sophisticated malware that it’s calling Reductor. The malware is able to compromise TLS traffic by infecting the computer with hacked TLS engine substituted on the fly, “marking” infected TLS handshakes by compromising the underlining random-number generator, and adding new digital certificates. The result is that the attacker can identify, intercept, and decrypt TLS traffic from the infected computer.

The Kaspersky Attribution Engine shows strong code similarities between this family and the COMPfun Trojan. Moreover, further research showed that the original COMpfun Trojan most probably is used as a downloader in one of the distribution schemes. Based on these similarities, we’re quite sure the new malware was developed by the COMPfun authors.

The COMpfun malware was initially documented by G-DATA in 2014. Although G-DATA didn’t identify which actor was using this malware, Kaspersky tentatively linked it to the Turla APT, based on the victimology. Our telemetry indicates that the current campaign using Reductor started at the end of April 2019 and remained active at the time of writing (August 2019). We identified targets in Russia and Belarus.

[…]

Turla has in the past shown many innovative ways to accomplish its goals, such as using hijacked satellite infrastructure. This time, if we’re right that Turla is the actor behind this new wave of attacks, then with Reductor it has implemented a very interesting way to mark a host’s encrypted TLS traffic by patching the browser without parsing network packets. The victimology for this new campaign aligns with previous Turla interests.

We didn’t observe any MitM functionality in the analyzed malware samples. However, Reductor is able to install digital certificates and mark the targets’ TLS traffic. It uses infected installers for initial infection through HTTP downloads from warez websites. The fact the original files on these sites are not infected also points to evidence of subsequent traffic manipulation.

The attribution chain from Reductor to COMPfun to Turla is thin. Speculation is that the attacker behind all of this is Russia.

Posted on October 10, 2019 at 1:49 PMView Comments

E-Mailing Private HTTPS Keys

I don’t know what to make of this story:

The email was sent on Tuesday by the CEO of Trustico, a UK-based reseller of TLS certificates issued by the browser-trusted certificate authorities Comodo and, until recently, Symantec. It was sent to Jeremy Rowley, an executive vice president at DigiCert, a certificate authority that acquired Symantec’s certificate issuance business after Symantec was caught flouting binding industry rules, prompting Google to distrust Symantec certificates in its Chrome browser. In communications earlier this month, Trustico notified DigiCert that 50,000 Symantec-issued certificates Trustico had resold should be mass revoked because of security concerns.

When Rowley asked for proof the certificates were compromised, the Trustico CEO emailed the private keys of 23,000 certificates, according to an account posted to a Mozilla security policy forum. The report produced a collective gasp among many security practitioners who said it demonstrated a shockingly cavalier treatment of the digital certificates that form one of the most basic foundations of website security.

Generally speaking, private keys for TLS certificates should never be archived by resellers, and, even in the rare cases where such storage is permissible, they should be tightly safeguarded. A CEO being able to attach the keys for 23,000 certificates to an email raises troubling concerns that those types of best practices weren’t followed.

I am croggled by the multiple layers of insecurity here.

BoingBoing post.

Posted on March 13, 2018 at 6:31 AMView Comments

How Signal Is Evading Censorship

Signal, the encrypted messaging app I prefer, is being blocked in both Egypt and the UAE. Recently, the Signal team developed a workaround: domain fronting.

Signal’s new anti-censorship feature uses a trick called “domain fronting,” Marlinspike explains. A country like Egypt, with only a few small internet service providers tightly controlled by the government, can block any direct request to a service on its blacklist. But clever services can circumvent that censorship by hiding their traffic inside of encrypted connections to a major internet service, like the content delivery networks (CDNs) that host content closer to users to speed up their online experience — or in Signal’s case, Google’s App Engine platform, designed to host apps on Google’s servers.

“Now when people in Egypt or the United Arab Emirates send a Signal message, it’ll look identical to something like a Google search,” Marlinspike says. “The idea is that using Signal will look like using Google; if you want to block Signal you’ll have to block Google.”

The trick works because Google’s App Engine allows developers to redirect traffic from Google.com to their own domain. Google’s use of TLS encryption means that contents of the traffic, including that redirect request, are hidden, and the internet service provider can see only that someone has connected to Google.com. That essentially turns Google into a proxy for Signal, bouncing its traffic and fooling the censors.

This isn’t a new trick (Tor uses it too, for example), but it does work.

Posted on December 28, 2016 at 6:20 AMView Comments

DROWN Attack

Earlier this week, we learned of yet another attack against SSL/TLS where an attacker can force people to use insecure algorithms. It’s called DROWN. Here’s a good news article on the attack, the technical paper describing the attack, and a very good technical blog post by Matthew Green.

As an aside, I am getting pretty annoyed at all the marketing surrounding vulnerabilities these days. Vulnerabilities do not need a catchy name, a dedicated website — even thought it’s a very good website — and a logo.

Posted on March 3, 2016 at 2:09 PMView Comments

Breaking Diffie-Hellman with Massive Precomputation (Again)

The Internet is abuzz with this blog post and paper, speculating that the NSA is breaking the Diffie-Hellman key-exchange protocol in the wild through massive precomputation.

I wrote about this at length in May when this paper was first made public. (The reason it’s news again is that the paper was just presented at the ACM Computer and Communications Security conference.)

What’s newly being talked about his how this works inside the NSA surveillance architecture. Nicholas Weaver explains:

To decrypt IPsec, a large number of wiretaps monitor for IKE (Internet Key Exchange) handshakes, the protocol that sets up a new IPsec encrypted connection. The handshakes are forwarded to a decryption oracle, a black box system that performs the magic. While this happens, the wiretaps also record all traffic in the associated IPsec connections.

After a period of time, this oracle either returns the private keys or says “i give up”. If the oracle provides the keys, the wiretap decrypts all the stored traffic and continues to decrypt the connection going forward.

[…]

This would also better match the security implications: just the fact that the NSA can decrypt a particular flow is a critical secret. Forwarding a small number of potentially-crackable flows to a central point better matches what is needed to maintain such secrecy.

Thus by performing the decryption in bulk at the wiretaps, complete with hardware acceleration to keep up with the number of encrypted streams, this architecture directly implies that the NSA can break a massive amount of IPsec traffic, a degree of success which implies a cryptanalysis breakthrough.

That last paragraph is Weaver explaining how this attack matches the NSA rhetoric about capabilities in some of their secret documents.

Now that this is out, I’m sure there are a lot of really upset people inside the NSA.

EDITED TO ADD (11/15): How to protect yourself.

Posted on October 16, 2015 at 6:19 AMView Comments

New RC4 Attack

New research: “All Your Biases Belong To Us: Breaking RC4 in WPA-TKIP and TLS,” by Mathy Vanhoef and Frank Piessens:

Abstract: We present new biases in RC4, break the Wi-Fi Protected Access Temporal Key Integrity Protocol (WPA-TKIP), and design a practical plaintext recovery attack against the Transport Layer Security (TLS) protocol. To empirically find new biases in the RC4 keystream we use statistical hypothesis tests. This reveals many new biases in the initial keystream bytes, as well as several new long-term biases. Our fixed-plaintext recovery algorithms are capable of using multiple types of biases, and return a list of plaintext candidates in decreasing likelihood.

To break WPA-TKIP we introduce a method to generate a large number of identical packets. This packet is decrypted by generating its plaintext candidate list, and using redundant packet structure to prune bad candidates. From the decrypted packet we derive the TKIP MIC key, which can be used to inject and decrypt packets. In practice the attack can be executed within an hour. We also attack TLS as used by HTTPS, where we show how to decrypt a secure cookie with a success rate of 94% using 9*227 ciphertexts. This is done by injecting known data around the cookie, abusing this using Mantin’s ABSAB bias, and brute-forcing the cookie by traversing the plaintext candidates. Using our traffic generation technique, we are able to execute the attack in merely 75 hours.

News articles.

We need to deprecate the algorithm already.

Posted on July 28, 2015 at 12:09 PMView Comments

The Logjam (and Another) Vulnerability against Diffie-Hellman Key Exchange

Logjam is a new attack against the Diffie-Hellman key-exchange protocol used in TLS. Basically:

The Logjam attack allows a man-in-the-middle attacker to downgrade vulnerable TLS connections to 512-bit export-grade cryptography. This allows the attacker to read and modify any data passed over the connection. The attack is reminiscent of the FREAK attack, but is due to a flaw in the TLS protocol rather than an implementation vulnerability, and attacks a Diffie-Hellman key exchange rather than an RSA key exchange. The attack affects any server that supports DHE_EXPORT ciphers, and affects all modern web browsers. 8.4% of the Top 1 Million domains were initially vulnerable.

Here’s the academic paper.

One of the problems with patching the vulnerability is that it breaks things:

On the plus side, the vulnerability has largely been patched thanks to consultation with tech companies like Google, and updates are available now or coming soon for Chrome, Firefox and other browsers. The bad news is that the fix rendered many sites unreachable, including the main website at the University of Michigan, which is home to many of the researchers that found the security hole.

This is a common problem with version downgrade attacks; patching them makes you incompatible with anyone who hasn’t patched. And it’s the vulnerability the media is focusing on.

Much more interesting is the other vulnerability that the researchers found:

Millions of HTTPS, SSH, and VPN servers all use the same prime numbers for Diffie-Hellman key exchange. Practitioners believed this was safe as long as new key exchange messages were generated for every connection. However, the first step in the number field sieve — the most efficient algorithm for breaking a Diffie-Hellman connection — is dependent only on this prime. After this first step, an attacker can quickly break individual connections.

The researchers believe the NSA has been using this attack:

We carried out this computation against the most common 512-bit prime used for TLS and demonstrate that the Logjam attack can be used to downgrade connections to 80% of TLS servers supporting DHE_EXPORT. We further estimate that an academic team can break a 768-bit prime and that a nation-state can break a 1024-bit prime. Breaking the single, most common 1024-bit prime used by web servers would allow passive eavesdropping on connections to 18% of the Top 1 Million HTTPS domains. A second prime would allow passive decryption of connections to 66% of VPN servers and 26% of SSH servers. A close reading of published NSA leaks shows that the agency’s attacks on VPNs are consistent with having achieved such a break.

Remember James Bamford’s 2012 comment about the NSA’s cryptanalytic capabilities:

According to another top official also involved with the program, the NSA made an enormous breakthrough several years ago in its ability to cryptanalyze, or break, unfathomably complex encryption systems employed by not only governments around the world but also many average computer users in the US. The upshot, according to this official: “Everybody’s a target; everybody with communication is a target.”

[…]

The breakthrough was enormous, says the former official, and soon afterward the agency pulled the shade down tight on the project, even within the intelligence community and Congress. “Only the chairman and vice chairman and the two staff directors of each intelligence committee were told about it,” he says. The reason? “They were thinking that this computing breakthrough was going to give them the ability to crack current public encryption.”

And remember Director of National Intelligence James Clapper’s introduction to the 2013 “Black Budget“:

Also, we are investing in groundbreaking cryptanalytic capabilities to defeat adversarial cryptography and exploit internet traffic.

It’s a reasonable guess that this is what both Bamford’s source and Clapper are talking about. It’s an attack that requires a lot of precomputation — just the sort of thing a national intelligence agency would go for.

But that requirement also speaks to its limitations. The NSA isn’t going to put this capability at collection points like Room 641A at AT&T’s San Francisco office: the precomputation table is too big, and the sensitivity of the capability is too high. More likely, an analyst identifies a target through some other means, and then looks for data by that target in databases like XKEYSCORE. Then he sends whatever ciphertext he finds to the Cryptanalysis and Exploitation Services (CES) group, which decrypts it if it can using this and other techniques.

Ross Anderson wrote about this earlier this month, almost certainly quoting Snowden:

As for crypto capabilities, a lot of stuff is decrypted automatically on ingest (e.g. using a “stolen cert”, presumably a private key obtained through hacking). Else the analyst sends the ciphertext to CES and they either decrypt it or say they can’t.

The analysts are instructed not to think about how this all works. This quote also applied to NSA employees:

Strict guidelines were laid down at the GCHQ complex in Cheltenham, Gloucestershire, on how to discuss projects relating to decryption. Analysts were instructed: “Do not ask about or speculate on sources or methods underpinning Bullrun.”

I remember the same instructions in documents I saw about the NSA’s CES.

Again, the NSA has put surveillance ahead of security. It never bothered to tell us that many of the “secure” encryption systems we were using were not secure. And we don’t know what other national intelligence agencies independently discovered and used this attack.

The good news is now that we know reusing prime numbers is a bad idea, we can stop doing it.

EDITED TO ADD: The DH precomputation easily lends itself to custom ASIC design, and is something that pipelines easily. Using BitCoin mining hardware as a rough comparison, this means a couple orders of magnitude speedup.

EDITED TO ADD (5/23): Good analysis of the cryptography.

EDITED TO ADD (5/24): Good explanation by Matthew Green.

Posted on May 21, 2015 at 6:30 AMView Comments

A New Free CA

Announcing Let’s Encrypt, a new free certificate authority. This is a joint project of EFF, Mozilla, Cisco, Akamai, and the University of Michigan.

This is an absolutely fantastic idea.

The anchor for any TLS-protected communication is a public-key certificate which demonstrates that the server you’re actually talking to is the server you intended to talk to. For many server operators, getting even a basic server certificate is just too much of a hassle. The application process can be confusing. It usually costs money. It’s tricky to install correctly. It’s a pain to update.

Let’s Encrypt is a new free certificate authority, built on a foundation of cooperation and openness, that lets everyone be up and running with basic server certificates for their domains through a simple one-click process.

[…]

The key principles behind Let’s Encrypt are:

  • Free: Anyone who owns a domain can get a certificate validated for that domain at zero cost.
  • Automatic: The entire enrollment process for certificates occurs painlessly during the server’s native installation or configuration process, while renewal occurs automatically in the background.
  • Secure: Let’s Encrypt will serve as a platform for implementing modern security techniques and best practices.
  • Transparent: All records of certificate issuance and revocation will be available to anyone who wishes to inspect them.
  • Open: The automated issuance and renewal protocol will be an open standard and as much of the software as possible will be open source.
  • Cooperative: Much like the underlying Internet protocols themselves, Let’s Encrypt is a joint effort to benefit the entire community, beyond the control of any one organization.

Slashdot thread. Hacker News thread.

EDITED TO ADD (11/19): Good post. And EFF’s blog post.

Posted on November 18, 2014 at 12:38 PMView Comments

Sidebar photo of Bruce Schneier by Joe MacInnis.