Blog: March 2018 Archives

Friday Squid Blogging: Giant Squid Stealing Food from Each Other

An interesting hunting strategy:

Off of northern Spain, giant squid often feed on schools of fish called blue whiting. The schools swim 400 meters or less below the surface, while the squid prefer to hang out around a mile deep. The squid must ascend to hunt, probably seizing fish from below with their tentacles, then descend again. In this scenario, a squid could save energy by pirating food from its neighbor rather than hunting its own fish, Guerra says: If the target squid has already carried its prey back to the depths to eat, the pirate could save itself a trip up to the shallow water. Staying below would also protect a pirate from predators such as dolphins and sperm whales that hang around the fish schools.

If a pirate happened to kill its victim, it would also reduce competition. The scientists think that's what happened with the Bares squid: Its tentacles were ripped off in the fight over food. "The victim, disoriented and wounded, could enter a warmer mass of water in which the efficiency of their blood decreases markedly," the authors write in a recent paper in the journal Ecology. "In this way, the victim, almost asphyxiated, would be at the mercy of the marine currents, being dragged toward the coast."

It's called "food piracy."

As usual, you can also use this squid post to talk about the security stories in the news that I haven't covered.

Read my blog posting guidelines here.

Posted on March 23, 2018 at 4:06 PM1 Comments

Zeynep Tufekci on Facebook and Cambridge Analytica

Zeynep Tufekci is particularly cogent about Facebook and Cambridge Analytica.

Several news outlets asked me to write about this issue. I didn't, because 1) my book manuscript is due on Monday (finally!), and 2) I knew Zeynep would say what I would say, only better.

Posted on March 23, 2018 at 2:21 PM12 Comments

GreyKey iPhone Unlocker

Some details about the iPhone unlocker from the US company Greyshift, with photos.

Little is known about Grayshift or its sales model at this point. We don't know whether sales are limited to US law enforcement, or if it is also selling in other parts of the world. Regardless of that, it's highly likely that these devices will ultimately end up in the hands of agents of an oppressive regime, whether directly from Grayshift or indirectly through the black market.

It's also entirely possible, based on the history of the IP-Box, that Grayshift devices will end up being available to anyone who wants them and can find a way to purchase them, perhaps by being reverse-engineered and reproduced by an enterprising hacker, then sold for a couple hundred bucks on eBay.

Forbes originally wrote about this, and I blogged that article.

Posted on March 23, 2018 at 6:28 AM11 Comments

Hijacking Computers for Cryptocurrency Mining

Interesting paper "A first look at browser-based cryptojacking":

Abstract: In this paper, we examine the recent trend towards in-browser mining of cryptocurrencies; in particular, the mining of Monero through Coinhive and similar code-bases. In this model, a user visiting a website will download a JavaScript code that executes client-side in her browser, mines a cryptocurrency, typically without her consent or knowledge, and pays out the seigniorage to the website. Websites may consciously employ this as an alternative or to supplement advertisement revenue, may offer premium content in exchange for mining, or may be unwittingly serving the code as a result of a breach (in which case the seigniorage is collected by the attacker). The cryptocurrency Monero is preferred seemingly for its unfriendliness to large-scale ASIC mining that would drive browser-based efforts out of the market, as well as for its purported privacy features. In this paper, we survey this landscape, conduct some measurements to establish its prevalence and profitability, outline an ethical framework for considering whether it should be classified as an attack or business opportunity, and make suggestions for the detection, mitigation and/or prevention of browser-based mining for non-consenting users.

Posted on March 21, 2018 at 6:27 AM32 Comments

Israeli Security Attacks AMD by Publishing Zero-Day Exploits

Last week, the Israeli security company CTS Labs published a series of exploits against AMD chips. The publication came with the flashy website, detailed whitepaper, cool vulnerability names -- RYZENFALL, MASTERKEY, FALLOUT, and CHIMERA -- and logos we've come to expect from these sorts of things. What's new is that the company only gave AMD a day's notice, which breaks with every norm about responsible disclosure. CTS Labs didn't release details of the exploits, only high-level descriptions of the vulnerabilities, but it is probably still enough for others to reproduce their results. This is incredibly irresponsible of the company.

Moreover, the vulnerabilities are kind of meh. Nicholas Weaver explains:

In order to use any of the four vulnerabilities, an attacker must already have almost complete control over the machine. For most purposes, if the attacker already has this access, we would generally say they've already won. But these days, modern computers at least attempt to protect against a rogue operating system by having separate secure subprocessors. CTS Labs discovered the vulnerabilities when they looked at AMD's implementation of the secure subprocessor to see if an attacker, having already taken control of the host operating system, could bypass these last lines of defense.

In a "Clarification," CTS Labs kind of agrees:

The vulnerabilities described in could give an attacker that has already gained initial foothold into one or more computers in the enterprise a significant advantage against IT and security teams.

The only thing the attacker would need after the initial local compromise is local admin privileges and an affected machine. To clarify misunderstandings -- there is no need for physical access, no digital signatures, no additional vulnerability to reflash an unsigned BIOS. Buy a computer from the store, run the exploits as admin -- and they will work (on the affected models as described on the site).

The weirdest thing about this story is that CTS Labs describes one of the vulnerabilities, Chimera, as a backdoor. Although it doesn't t come out and say that this was deliberately planted by someone, it does make the point that the chips were designed in Taiwan. This is an incredible accusation, and honestly needs more evidence before we can evaluate it.

The upshot of all of this is that CTS Labs played this for maximum publicity: over-hyping its results and minimizing AMD's ability to respond. And it may have an ulterior motive:

But CTS's website touting AMD's flaws also contained a disclaimer that threw some shadows on the company's motives: "Although we have a good faith belief in our analysis and believe it to be objective and unbiased, you are advised that we may have, either directly or indirectly, an economic interest in the performance of the securities of the companies whose products are the subject of our reports," reads one line. WIRED asked in a follow-up email to CTS whether the company holds any financial positions designed to profit from the release of its AMD research specifically. CTS didn't respond.

We all need to demand better behavior from security researchers. I know that any publicity is good publicity, but I am pleased to see the stories critical of CTS Labs outnumbering the stories praising it.

EDITED TO ADD (3/21): AMD responds:

AMD's response today agrees that all four bug families are real and are found in the various components identified by CTS. The company says that it is developing firmware updates for the three PSP flaws. These fixes, to be made available in "coming weeks," will be installed through system firmware updates. The firmware updates will also mitigate, in some unspecified way, the Chimera issue, with AMD saying that it's working with ASMedia, the third-party hardware company that developed Promontory for AMD, to develop suitable protections. In its report, CTS wrote that, while one CTS attack vector was a firmware bug (and hence in principle correctable), the other was a hardware flaw. If true, there may be no effective way of solving it.

Response here.

Posted on March 19, 2018 at 6:27 AM44 Comments

Friday Squid Blogging: New Squid Species Discovered in Australia

A new species of pygmy squid was discovered in Western Australia. It's pretty cute.

As usual, you can also use this squid post to talk about the security stories in the news that I haven't covered.

Read my blog posting guidelines here.

Posted on March 16, 2018 at 4:10 PM168 Comments

Artificial Intelligence and the Attack/Defense Balance

Artificial intelligence technologies have the potential to upend the longstanding advantage that attack has over defense on the Internet. This has to do with the relative strengths and weaknesses of people and computers, how those all interplay in Internet security, and where AI technologies might change things.

You can divide Internet security tasks into two sets: what humans do well and what computers do well. Traditionally, computers excel at speed, scale, and scope. They can launch attacks in milliseconds and infect millions of computers. They can scan computer code to look for particular kinds of vulnerabilities, and data packets to identify particular kinds of attacks.

Humans, conversely, excel at thinking and reasoning. They can look at the data and distinguish a real attack from a false alarm, understand the attack as it's happening, and respond to it. They can find new sorts of vulnerabilities in systems. Humans are creative and adaptive, and can understand context.

Computers -- so far, at least -- are bad at what humans do well. They're not creative or adaptive. They don't understand context. They can behave irrationally because of those things.

Humans are slow, and get bored at repetitive tasks. They're terrible at big data analysis. They use cognitive shortcuts, and can only keep a few data points in their head at a time. They can also behave irrationally because of those things.

AI will allow computers to take over Internet security tasks from humans, and then do them faster and at scale. Here are possible AI capabilities:

  • Discovering new vulnerabilities­ -- and, more importantly, new types of vulnerabilities­ in systems, both by the offense to exploit and by the defense to patch, and then automatically exploiting or patching them.
  • Reacting and adapting to an adversary's actions, again both on the offense and defense sides. This includes reasoning about those actions and what they mean in the context of the attack and the environment.
  • Abstracting lessons from individual incidents, generalizing them across systems and networks, and applying those lessons to increase attack and defense effectiveness elsewhere.
  • Identifying strategic and tactical trends from large datasets and using those trends to adapt attack and defense tactics.

That's an incomplete list. I don't think anyone can predict what AI technologies will be capable of. But it's not unreasonable to look at what humans do today and imagine a future where AIs are doing the same things, only at computer speeds, scale, and scope.

Both attack and defense will benefit from AI technologies, but I believe that AI has the capability to tip the scales more toward defense. There will be better offensive and defensive AI techniques. But here's the thing: defense is currently in a worse position than offense precisely because of the human components. Present-day attacks pit the relative advantages of computers and humans against the relative weaknesses of computers and humans. Computers moving into what are traditionally human areas will rebalance that equation.

Roy Amara famously said that we overestimate the short-term effects of new technologies, but underestimate their long-term effects. AI is notoriously hard to predict, so many of the details I speculate about are likely to be wrong­ -- and AI is likely to introduce new asymmetries that we can't foresee. But AI is the most promising technology I've seen for bringing defense up to par with offense. For Internet security, that will change everything.

This essay previously appeared in the March/April 2018 issue of IEEE Security & Privacy.

Posted on March 15, 2018 at 6:16 AM43 Comments

The 600+ Companies PayPal Shares Your Data With

One of the effects of GDPR -- the new EU General Data Protection Regulation -- is that we're all going to be learning a lot more about who collects our data and what they do with it. Consider PayPal, that just released a list of over 600 companies they share customer data with. Here's a good visualization of that data.

Is 600 companies unusual? Is it more than average? Less? We'll soon know.

Posted on March 14, 2018 at 6:24 AM39 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 AM43 Comments

Two New Papers on the Encryption Debate

Seems like everyone is writing about encryption and backdoors this season.

I recently blogged about the new National Academies report on the same topic.

Here's a review of the National Academies report, and another of the East West Institute's report.

EDITED TO ADD (3/8): Commentary on the National Academies study by the EFF.

Posted on March 12, 2018 at 6:27 AM21 Comments

Friday Squid Blogging: Interesting Interview

Here's an hour-long audio interview with squid scientist Sarah McAnulty.

As usual, you can also use this squid post to talk about the security stories in the news that I haven't covered.

Read my blog posting guidelines here.

Posted on March 9, 2018 at 4:22 PM179 Comments

OURSA Conference

Responding to the lack of diversity at the RSA Conference, a group of security experts have announced a competing one-day conference: OUR Security Advocates, or OURSA. It's in San Francisco, and it's during RSA, so you can attend both.

Posted on March 9, 2018 at 6:24 AM35 Comments

New DDoS Reflection-Attack Variant

This is worrisome:

DDoS vandals have long intensified their attacks by sending a small number of specially designed data packets to publicly available services. The services then unwittingly respond by sending a much larger number of unwanted packets to a target. The best known vectors for these DDoS amplification attacks are poorly secured domain name system resolution servers, which magnify volumes by as much as 50 fold, and network time protocol, which increases volumes by about 58 times.

On Tuesday, researchers reported attackers are abusing a previously obscure method that delivers attacks 51,000 times their original size, making it by far the biggest amplification method ever used in the wild. The vector this time is memcached, a database caching system for speeding up websites and networks. Over the past week, attackers have started abusing it to deliver DDoSes with volumes of 500 gigabits per second and bigger, DDoS mitigation service Arbor Networks reported in a blog post.

Cloudflare blog post. BoingBoing post.

EDITED TO ADD (3/9): Brian Krebs covered this.

Posted on March 7, 2018 at 6:23 AM13 Comments

Security Vulnerabilities in Smart Contracts

Interesting research: "Finding The Greedy, Prodigal, and Suicidal Contracts at Scale":

Abstract: Smart contracts -- stateful executable objects hosted on blockchains like Ethereum -- carry billions of dollars worth of coins and cannot be updated once deployed. We present a new systematic characterization of a class of trace vulnerabilities, which result from analyzing multiple invocations of a contract over its lifetime. We focus attention on three example properties of such trace vulnerabilities: finding contracts that either lock funds indefinitely, leak them carelessly to arbitrary users, or can be killed by anyone. We implemented MAIAN, the first tool for precisely specifying and reasoning about trace properties, which employs inter-procedural symbolic analysis and concrete validator for exhibiting real exploits. Our analysis of nearly one million contracts flags 34,200 (2,365 distinct) contracts vulnerable, in 10 seconds per contract. On a subset of 3,759 contracts which we sampled for concrete validation and manual analysis, we reproduce real exploits at a true positive rate of 89%, yielding exploits for 3,686 contracts. Our tool finds exploits for the infamous Parity bug that indirectly locked 200 million dollars worth in Ether, which previous analyses failed to capture.

Posted on March 6, 2018 at 6:18 AM24 Comments

Intimate Partner Threat

Princeton's Karen Levy has a good article computer security and the intimate partner threat:

When you learn that your privacy has been compromised, the common advice is to prevent additional access -- delete your insecure account, open a new one, change your password. This advice is such standard protocol for personal security that it's almost a no-brainer. But in abusive romantic relationships, disconnection can be extremely fraught. For one, it can put the victim at risk of physical harm: If abusers expect digital access and that access is suddenly closed off, it can lead them to become more violent or intrusive in other ways. It may seem cathartic to delete abusive material, like alarming text messages -- but if you don't preserve that kind of evidence, it can make prosecution more difficult. And closing some kinds of accounts, like social networks, to hide from a determined abuser can cut off social support that survivors desperately need. In some cases, maintaining a digital connection to the abuser may even be legally required (for instance, if the abuser and survivor share joint custody of children).

Threats from intimate partners also change the nature of what it means to be authenticated online. In most contexts, access credentials­ -- like passwords and security questions -- are intended to insulate your accounts against access from an adversary. But those mechanisms are often completely ineffective for security in intimate contexts: The abuser can compel disclosure of your password through threats of violence and has access to your devices because you're in the same physical space. In many cases, the abuser might even own your phone -- or might have access to your communications data because you share a family plan. Things like security questions are unlikely to be effective tools for protecting your security, because the abuser knows or can guess at intimate details about your life -- where you were born, what your first job was, the name of your pet.

Posted on March 5, 2018 at 11:13 AM37 Comments

Extracting Secrets from Machine Learning Systems

This is fascinating research about how the underlying training data for a machine-learning system can be inadvertently exposed. Basically, if a machine-learning system trains on a dataset that contains secret information, in some cases an attacker can query the system to extract that secret information. My guess is that there is a lot more research to be done here.

EDITED TO ADD (3/9): Some interesting links on the subject.

Posted on March 5, 2018 at 5:20 AM10 Comments

Friday Squid Blogging: Searching for Humboldt Squid with Electronic Bait

Video and short commentary.

As usual, you can also use this squid post to talk about the security stories in the news that I haven't covered.

Read my blog posting guidelines here.

Posted on March 2, 2018 at 4:11 PM128 Comments

Malware from Space

Since you don't have enough to worry about, here's a paper postulating that space aliens could send us malware capable of destroying humanity.

Abstract: A complex message from space may require the use of computers to display, analyze and understand. Such a message cannot be decontaminated with certainty, and technical risks remain which can pose an existential threat. Complex messages would need to be destroyed in the risk averse case.

I think we're more likely to be enslaved by malicious AIs.

Posted on March 2, 2018 at 6:13 AM87 Comments

Russians Hacked the Olympics

Two weeks ago, I blogged about the myriad of hacking threats against the Olympics. Last week, the Washington Post reported that Russia hacked the Olympics network and tried to cast the blame on North Korea.

Of course, the evidence is classified, so there's no way to verify this claim. And while the article speculates that the hacks were a retaliation for Russia being banned due to doping, that doesn't ring true to me. If they tried to blame North Korea, it's more likely that they're trying to disrupt something between North Korea, South Korea, and the US. But I don't know.

Posted on March 1, 2018 at 6:47 AM101 Comments

Photo of Bruce Schneier by Per Ervland.

Schneier on Security is a personal website. Opinions expressed are not necessarily those of IBM Resilient.