All posts by Bruce Schneier

Recovering Passwords by Measuring Residual Heat

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/10/recovering-passwords-by-measuring-residual-heat.html

Researchers have used thermal cameras and ML guessing techniques to recover passwords from measuring the residual heat left by fingers on keyboards. From the abstract:

We detail the implementation of ThermoSecure and make a dataset of 1,500 thermal images of keyboards with heat traces resulting from input publicly available. Our first study shows that ThermoSecure successfully attacks 6-symbol, 8-symbol, 12-symbol, and 16-symbol passwords with an average accuracy of 92%, 80%, 71%, and 55% respectively, and even higher accuracy when thermal images are taken within 30 seconds. We found that typing behavior significantly impacts vulnerability to thermal attacks, where hunt-and-peck typists are more vulnerable than fast typists (92% vs 83% thermal attack success if performed within 30 seconds). The second study showed that the keycaps material has a statistically significant effect on the effectiveness of thermal attacks: ABS keycaps retain the thermal trace of users presses for a longer period of time, making them more vulnerable to thermal attacks, with a 52% average attack accuracy compared to 14% for keyboards with PBT keycaps.

“ABS” is Acrylonitrile Butadiene Styrene, which some keys are made of. Others are made of Polybutylene Terephthalate (PBT). PBT keys are less vulnerable.

But, honestly, if someone can train a camera at your keyboard, you have bigger problems.

News article.

Inserting a Backdoor into a Machine-Learning System

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/10/inserting-a-backdoor-into-a-machine-learning-system.html

Interesting research: “ImpNet: Imperceptible and blackbox-undetectable backdoors in compiled neural networks, by Tim Clifford, Ilia Shumailov, Yiren Zhao, Ross Anderson, and Robert Mullins:

Abstract: Early backdoor attacks against machine learning set off an arms race in attack and defence development. Defences have since appeared demonstrating some ability to detect backdoors in models or even remove them. These defences work by inspecting the training data, the model, or the integrity of the training procedure. In this work, we show that backdoors can be added during compilation, circumventing any safeguards in the data preparation and model training stages. As an illustration, the attacker can insert weight-based backdoors during the hardware compilation step that will not be detected by any training or data-preparation process. Next, we demonstrate that some backdoors, such as ImpNet, can only be reliably detected at the stage where they are inserted and removing them anywhere else presents a significant challenge. We conclude that machine-learning model security requires assurance of provenance along the entire technical pipeline, including the data, model architecture, compiler, and hardware specification.

Ross Anderson explains the significance:

The trick is for the compiler to recognise what sort of model it’s compiling—whether it’s processing images or text, for example—and then devising trigger mechanisms for such models that are sufficiently covert and general. The takeaway message is that for a machine-learning model to be trustworthy, you need to assure the provenance of the whole chain: the model itself, the software tools used to compile it, the training data, the order in which the data are batched and presented—in short, everything.

Complex Impersonation Story

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/10/complex-impersonation-story.html

This is a story of one piece of what is probably a complex employment scam. Basically, real programmers are having their resumes copied and co-opted by scammers, who apply for jobs (or, I suppose, get recruited from various job sites), then hire other people with Western looks and language skills are to impersonate those first people on Zoom job interviews. Presumably, sometimes the scammers get hired and…I suppose…collect paychecks for a while until they get found out and fired. But that requires a bunch of banking fraud as well, so I don’t know.

EDITED TO ADD (10/11): Brian Krebs writes about fake LinkedIn profiles, which is probably another facet of this fraud system. Someone needs to unravel all of the threads.

Friday Squid Blogging: Emotional Support Squid

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/10/friday-squid-blogging-emotional-support-squid.html

The Monterey Bay Aquarium has a video—”2 Hours Of Squid To Relax/Study/Work To“—with 2.4 million views.

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.

Spyware Maker Intellexa Sued by Journalist

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/10/spyware-maker-intellexa-sued-by-journalist.html

The Greek journalist Thanasis Koukakis was spied on by his own government, with a commercial spyware product called “Predator.” That product is sold by a company in North Macedonia called Cytrox, which is in turn owned by an Israeli company called Intellexa.

Koukakis is suing Intellexa.

The lawsuit filed by Koukakis takes aim at Intellexa and its executive, alleging a criminal breach of privacy and communication laws, reports Haaretz. The founder of Intellexa, a former Israeli intelligence commander named Taj Dilian, is listed as one of the defendants in the suit, as is another shareholder, Sara Hemo, and the firm itself. The objective of the suit, Koukakis says, is to spur an investigation to determine whether a criminal indictment should be brought against the defendants.

Why does it always seem to be Israel? The world would be a much safer place if that government stopped this cyberweapons arms trade from inside its borders.

October Is Cybersecurity Awareness Month

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/10/october-is-cybersecurity-awareness-month.html

For the past nineteen years, October has been Cybersecurity Awareness Month here in the US, and that event that has always been part advice and part ridicule. I tend to fall on the apathy end of the spectrum; I don’t think I’ve ever mentioned it before. But the memes can be funny.

Here’s a decent rundown of some of the chatter.

NSA Employee Charged with Espionage

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/10/nsa-employee-charged-with-espionage.html

An ex-NSA employee has been charged with trying to sell classified data to the Russians (but instead actually talking to an undercover FBI agent).

It’s a weird story, and the FBI affidavit raises more questions than it answers. The employee only worked for the NSA for three weeks—which is weird in itself. I can’t figure out how he linked up with the undercover FBI agent. It’s not clear how much of this was the employee’s idea, and whether he was goaded by the FBI agent. Still, hooray for not leaking NSA secrets to the Russians. (And, almost ten years after Snowden, do we still have this much trouble vetting people before giving them security clearances?)

Mr. Dalke, who had already left the N.S.A. but told the agent that he still worked there on a temporary assignment, then revealed that had taken “highly sensitive information” related to foreign targeting of U.S. systems and information on cyber operations, the prosecutors said. He offered the information in exchange for cryptocurrency and said he was in “financial need.” Court records show he had nearly $84,000 in debt between student loans and credit cards.

EDITED TO ADD (10/5): Marcy Wheeler notes that the FBI seems to be sitting on some common recruitment point, and collecting potential Russian spies.

Detecting Deepfake Audio by Modeling the Human Acoustic Tract

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/10/detecting-deepfake-audio-by-modeling-the-human-acoustic-tract.html

This is interesting research:

In this paper, we develop a new mechanism for detecting audio deepfakes using techniques from the field of articulatory phonetics. Specifically, we apply fluid dynamics to estimate the arrangement of the human vocal tract during speech generation and show that deepfakes often model impossible or highly-unlikely anatomical arrangements. When parameterized to achieve 99.9% precision, our detection mechanism achieves a recall of 99.5%, correctly identifying all but one deepfake sample in our dataset.

From an article by two of the researchers:

The first step in differentiating speech produced by humans from speech generated by deepfakes is understanding how to acoustically model the vocal tract. Luckily scientists have techniques to estimate what someone—or some being such as a dinosaur—would sound like based on anatomical measurements of its vocal tract.

We did the reverse. By inverting many of these same techniques, we were able to extract an approximation of a speaker’s vocal tract during a segment of speech. This allowed us to effectively peer into the anatomy of the speaker who created the audio sample.

From here, we hypothesized that deepfake audio samples would fail to be constrained by the same anatomical limitations humans have. In other words, the analysis of deepfaked audio samples simulated vocal tract shapes that do not exist in people.

Our testing results not only confirmed our hypothesis but revealed something interesting. When extracting vocal tract estimations from deepfake audio, we found that the estimations were often comically incorrect. For instance, it was common for deepfake audio to result in vocal tracts with the same relative diameter and consistency as a drinking straw, in contrast to human vocal tracts, which are much wider and more variable in shape.

This is, of course, not the last word. Deepfake generators will figure out how to use these techniques to create harder-to-detect fake voices. And the deepfake detectors will figure out another, better, detection technique. And the arms race will continue.

Slashdot thread.

Security Vulnerabilities in Covert CIA Websites

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/09/security-vulnerabilities-in-covert-cia-websites.html

Back in 2018, we learned that covert system of websites that the CIA used for communications was compromised by—at least—China and Iran, and that the blunder caused a bunch of arrests, imprisonments, and executions. We’re now learning that the CIA is still “using an irresponsibly secured system for asset communication.”

Citizen Lab did the research:

Using only a single website, as well as publicly available material such as historical internet scanning results and the Internet Archive’s Wayback Machine, we identified a network of 885 websites and have high confidence that the United States (US) Central Intelligence Agency (CIA) used these sites for covert communication.

The websites included similar Java, JavaScript, Adobe Flash, and CGI artifacts that implemented or apparently loaded covert communications apps. In addition, blocks of sequential IP addresses registered to apparently fictitious US companies were used to host some of the websites. All of these flaws would have facilitated discovery by hostile parties.

[…]

The bulk of the websites that we discovered were active at various periods between 2004 and 2013. We do not believe that the CIA has recently used this communications infrastructure. Nevertheless, a subset of the websites are linked to individuals who may be former and possibly still active intelligence community employees or assets:

  • Several are currently abroad
  • Another left mainland China in the timeframe of the Chinese crackdown
  • Another was subsequently employed by the US State Department
  • Another now works at a foreign intelligence contractor

Citizen Lab is not publishing details, of course.

When I was a kid, I thought a lot about being a spy. And this, right here, was the one thing I worried about. It didn’t matter how clever and resourceful I was. If my handlers were incompetent, I was dead.

Another news article.

EDITED TO ADD (10/2): Slashdot thread.

Differences in App Security/Privacy Based on Country

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/09/differences-in-app-security-privacy-based-on-country.html

Depending on where you are when you download your Android apps, it might collect more or less data about you.

The apps we downloaded from Google Play also showed differences based on country in their security and privacy capabilities. One hundred twenty-seven apps varied in what the apps were allowed to access on users’ mobile phones, 49 of which had additional permissions deemed “dangerous” by Google. Apps in Bahrain, Tunisia and Canada requested the most additional dangerous permissions.

Three VPN apps enable clear text communication in some countries, which allows unauthorized access to users’ communications. One hundred and eighteen apps varied in the number of ad trackers included in an app in some countries, with the categories Games, Entertainment and Social, with Iran and Ukraine having the most increases in the number of ad trackers compared to the baseline number common to all countries.

One hundred and three apps have differences based on country in their privacy policies. Users in countries not covered by data protection regulations, such as GDPR in the EU and the California Consumer Privacy Act in the U.S., are at higher privacy risk. For instance, 71 apps available from Google Play have clauses to comply with GDPR only in the EU and CCPA only in the U.S. Twenty-eight apps that use dangerous permissions make no mention of it, despite Google’s policy requiring them to do so.

Research paper: “A Large-scale Investigation into Geodifferences in Mobile Apps“:

Abstract: Recent studies on the web ecosystem have been raising alarms on the increasing geodifferences in access to Internet content and services due to Internet censorship and geoblocking. However, geodifferences in the mobile app ecosystem have received limited attention, even though apps are central to how mobile users communicate and consume Internet content. We present the first large-scale measurement study of geodifferences in the mobile app ecosystem. We design a semi-automatic, parallel measurement testbed that we use to collect 5,684 popular apps from Google Play in 26 countries. In all, we collected 117,233 apk files and 112,607 privacy policies for those apps. Our results show high amounts of geoblocking with 3,672 apps geoblocked in at least one of our countries. While our data corroborates anecdotal evidence of takedowns due to government requests, unlike common perception, we find that blocking by developers is significantly higher than takedowns in all our countries, and has the most influence on geoblocking in the mobile app ecosystem. We also find instances of developers releasing different app versions to different countries, some with weaker security settings or privacy disclosures that expose users to higher security and privacy risks. We provide recommendations for app market proprietors to address the issues discovered.

Cold War Bugging of Soviet Facilities

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/09/cold-war-bugging-of-soviet-facilities.html

Found documents in Poland detail US spying operations against the former Soviet Union.

The file details a number of bugs found at Soviet diplomatic facilities in Washington, D.C., New York, and San Francisco, as well as in a Russian government-owned vacation compound, apartments used by Russia personnel, and even Russian diplomats’ cars. And the bugs were everywhere: encased in plaster in an apartment closet; behind electrical and television outlets; bored into concrete bricks and threaded into window frames; inside wooden beams and baseboards and stashed within a building’s foundation itself; surreptitiously attached to security cameras; wired into ceiling panels and walls; and secretly implanted into the backseat of cars and in their window panels, instrument panels, and dashboards. It’s an impressive—­ and impressively thorough—­ effort by U.S. counterspies.

We have long read about sophisticated Russian spying operations—bugging the Moscow embassy, bugging Selectric typewriters in the Moscow embassy, bugging the new Moscow embassy. These are the first details I’ve read about the US bugging the Russians’ embassy.

New Report on IoT Security

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/09/new-report-on-iot-security.html

The Atlantic Council has published a report on securing the Internet of Things: “Security in the Billions: Toward a Multinational Strategy to Better Secure the IoT Ecosystem.” The report examines the regulatory approaches taken by four countries—the US, the UK, Australia, and Singapore—to secure home, medical, and networking/telecommunications devices. The report recommends that regulators should 1) enforce minimum security standards for manufacturers of IoT devices, 2) incentivize higher levels of security through public contracting, and 3) try to align IoT standards internationally (for example, international guidance on handling connected devices that stop receiving security updates).

This report looks to existing security initiatives as much as possible—both to leverage existing work and to avoid counterproductively suggesting an entirely new approach to IoT security—while recommending changes and introducing more cohesion and coordination to regulatory approaches to IoT cybersecurity. It walks through the current state of risk in the ecosystem, analyzes challenges with the current policy model, and describes a synthesized IoT security framework. The report then lays out nine recommendations for government and industry actors to enhance IoT security, broken into three recommendation sets: setting a baseline of minimally acceptable security (or “Tier 1”), incentivizing above the baseline (or “Tier 2” and above), and pursuing international alignment on standards and implementation across the entire IoT product lifecycle (from design to sunsetting). It also includes implementation guidance for the United States, Australia, UK, and Singapore, providing a clearer roadmap for countries to operationalize the recommendations in their specific jurisdictions—and push towards a stronger, more cohesive multinational approach to securing the IoT worldwide.

Note: One of the authors of this report was a student of mine at Harvard Kennedy School, and did this work with the Atlantic Council under my supervision.

Leaking Passwords through the Spellchecker

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/09/leaking-passwords-through-the-spellchecker.html

Sometimes browser spellcheckers leak passwords:

When using major web browsers like Chrome and Edge, your form data is transmitted to Google and Microsoft, respectively, should enhanced spellcheck features be enabled.

Depending on the website you visit, the form data may itself include PII­—including but not limited to Social Security Numbers (SSNs)/Social Insurance Numbers (SINs), name, address, email, date of birth (DOB), contact information, bank and payment information, and so on.

The solution is to only use the spellchecker options that keep the data on your computer—and don’t send it into the cloud.

Friday Squid Blogging: Another Giant Squid Washes Up on New Zealand Beach

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/09/friday-squid-blogging-another-giant-squid-washes-up-on-new-zealand-beach.html

This one has chewed-up tentacles.

(Note that this is a different squid than the one that recently washed up on a South African beach.)

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.

Leaking Screen Information on Zoom Calls through Reflections in Eyeglasses

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/09/leaking-screen-information-on-zoom-calls-through-reflections-in-eyeglasses.html

Okay, it’s an obscure threat. But people are researching it:

Our models and experimental results in a controlled lab setting show it is possible to reconstruct and recognize with over 75 percent accuracy on-screen texts that have heights as small as 10 mm with a 720p webcam.” That corresponds to 28 pt, a font size commonly used for headings and small headlines.

[…]

Being able to read reflected headline-size text isn’t quite the privacy and security problem of being able to read smaller 9 to 12 pt fonts. But this technique is expected to provide access to smaller font sizes as high-resolution webcams become more common.

“We found future 4k cameras will be able to peek at most header texts on almost all websites and some text documents,” said Long.

[…]

A variety of factors can affect the legibility of text reflected in a video conference participant’s glasses. These include reflectance based on the meeting participant’s skin color, environmental light intensity, screen brightness, the contrast of the text with the webpage or application background, and the characteristics of eyeglass lenses. Consequently, not every glasses-wearing person will necessarily provide adversaries with reflected screen sharing.

With regard to potential mitigations, the boffins say that Zoom already provides a video filter in its Background and Effects settings menu that consists of reflection-blocking opaque cartoon glasses. Skype and Google Meet lack that defense.

Research paper.

Automatic Cheating Detection in Human Racing

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/09/automatic-cheating-detection-in-human-racing.html

This is a fascinating glimpse of the future of automatic cheating detection in sports:

Maybe you heard about the truly insane false-start controversy in track and field? Devon Allen—a wide receiver for the Philadelphia Eagles—was disqualified from the 110-meter hurdles at the World Athletics Championships a few weeks ago for a false start.

Here’s the problem: You can’t see the false start. Nobody can see the false start. By sight, Allen most definitely does not leave before the gun.

But here’s the thing: World Athletics has determined that it is not possible for someone to push off the block within a tenth of a second of the gun without false starting. They have science that shows it is beyond human capabilities to react that fast. Of course there are those (I’m among them) who would tell you that’s nonsense, that’s pseudoscience, there’s no way that they can limit human capabilities like that. There is science that shows it is humanly impossible to hit a fastball. There was once science that showed human beings could not run a four-minute mile.

Besides, do you know what Devon Allen’s reaction time was? It was 0.99 seconds. One thousandth of a second too fast, according to World Athletics’ science. They’re THAT sure that .01 seconds—and EXACTLY .01 seconds—is the limit of human possibilities that they will disqualify an athlete who has trained his whole life for this moment because he reacted one thousandth of a second faster than they think possible?

We in the computer world are used to this sort of thing. “The computer is always right,” even when it’s obviously wrong. But now computers are leaving the world of keyboards and screens, and this sort of thing will become more pervasive. In sports, computer systems are used to detect when a ball is out of bounds in tennis and other games and when a pitch is a strike in baseball. I’m sure there’s more—are computers detecting first downs in football?—but I’m not enough of a sports person to know them.

Credit Card Fraud That Bypasses 2FA

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/09/credit-card-fraud-that-bypasses-2fa.html

Someone in the UK is stealing smartphones and credit cards from people who have stored them in gym lockers, and is using the two items in combination to commit fraud:

Phones, of course, can be made inaccessible with the use of passwords and face or fingerprint unlocking. And bank cards can be stopped.

But the thief has a method which circumnavigates those basic safety protocols.

Once they have the phone and the card, they register the card on the relevant bank’s app on their own phone or computer. Since it is the first time that card will have been used on the new device, a one-off security passcode is demanded.

That verification passcode is sent by the bank to the stolen phone. The code flashes up on the locked screen of the stolen phone, leaving the thief to tap it into their own device. Once accepted, they have control of the bank account. They can transfer money or buy goods, or change access to the account.

Large-Scale Collection of Cell Phone Data at US Borders

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2022/09/large-scale-collection-of-cell-phone-data-at-us-borders.html

The Washington Post is reporting that the US Customs and Border Protection agency is seizing and copying cell phone, tablet, and computer data from “as many as” 10,000 phones per year, including an unspecified number of American citizens. This is done without a warrant, because “…courts have long granted an exception to border authorities, allowing them to search people’s devices without a warrant or suspicion of a crime.”

CBP’s inspection of people’s phones, laptops, tablets and other electronic devices as they enter the country has long been a controversial practice that the agency has defended as a low-impact way to pursue possible security threats and determine an individual’s “intentions upon entry” into the U.S. But the revelation that thousands of agents have access to a searchable database without public oversight is a new development in what privacy advocates and some lawmakers warn could be an infringement of Americans’ Fourth Amendment rights against unreasonable searches and seizures.

[…]

CBP conducted roughly 37,000 searches of travelers’ devices in the 12 months ending in October 2021, according to agency data, and more than 179 million people traveled that year through U.S. ports of entry.

More articles. Slashdot thread.