Silenda Flash. Flash drive with physical data destruction function.

Man

Professional
Messages
3,059
Reaction score
585
Points
113
Can you be 100% sure that your protection methods (or lack thereof) will help you in a critical situation? There is only one effective way - to destroy the information.

Hello, Anonymous!
Where do you store your gigabytes? On a flash drive in a secret place? In a cryptocontainer? Or maybe in the cloud or on Yandex disk? Can you be 100% sure that your protection methods (or lack thereof) will help you in a critical situation?

There is only one effective way - to destroy.

Yes, my friend, today we will talk about devices with an automatic destruction function.
Today we will talk about a new generation device that allows not only to encrypt, but also to destroy information by pressing a button on the flash drive body. And also to activate the destruction mechanism by giving a deliberately false password.

Meet Silenda Flash


9cb38a0e-55e6-4f70-b7d2-0f1d00dee7d0.png


This is not a commissioned article. During the testing of this device, I tested the hardware and software for almost two months, because I, like no one else, understand and know from my own experience how important it is to be sure of the reliability of storing your information and where it is stored. There are no "buts" here. Therefore, this article contains only facts.

What do the manufacturers write?​

According to the manufacturers, Silenda Flash is a 64GB flash drive with self-destruction and encryption capabilities. The developers claim the following functionality (information taken from the developers' website ):

Physical destruction of data​

Favorably different from its competitors, Silenda Flash guarantees physical destruction of data, while other devices designed to ensure data security simply "erase" information by rewriting. Silenda Flash is equipped with a built-in battery, and on the body of the carrier there is a button, pressing which physically destroys the memory chip itself due to a directed electric discharge with a voltage of up to 74 volts, which guarantees the impossibility of data recovery.

Data encryption​

At the user's request, access to information on Silenda Flash is possible only with a password, which is set when encrypting data on the drive. Thus, the information is protected even if it falls into the wrong hands.

Possibility of setting a deliberately false password, leading to destruction of data​

The user can set a second password, which, when entered, instantly physically deletes data from the storage device.

Silenda Flash as a portable workstation​

The device can be used as a remote workstation. The available volume is quite enough to install Windows or Tails with a small cryptocontainer.

What is it really?​

I want to say right away that this is the second version of the device. The first one was tested back in 2015 and written about by the guys from the magazine "Hacker", a resource that I have respected since early childhood. The article from "Hacker" showed me that the manufacturer has sufficient experience in developing information security devices. And I was eager to see and check for myself what it is.

The device is made in the form of a flash drive, in a black plastic case with a button and an LED. Inside there is a printed circuit board of fairly good quality with a micro-SD card soldered to it. The chip markings are erased, but the manufacturer reported that a USB hub from MJ Microchip is used. The micro-SD card itself and the Arduino control module are connected directly to the controller (the manufacturer refused to provide the markings). The battery is located above. The physical destruction mechanism is activated by discharging a capacitor with a peak voltage of up to 74 volts, depending on the battery charge. This is quite enough to physically destroy information on the media. However, physical protection against opening the case is not implemented. This means that if you did not manage to press the button in time, everything will depend only on the strength of your password, because the battery can simply be disconnected.

14d81ca4-9faf-4808-a639-2630c3570b83.png


Shall we connect?​

When connected to a computer, the device is identified as a regular 64GB flash drive. And… As a HID device, that is, Human Interface Device, or in Russian, an input device, with the manufacturer ID Van Ooijen Technische Informatica Datakiller. What is it? HID devices are, for example, mice or keyboards…

Here we need to explain a little. The USB standard does not protect against some devices imitating others. During the initialization process, the microcontroller informs the host, along with other service information, the classes to which the device belongs. The host loads the necessary driver and works with the device based on its class and this data. In fact, the flash drive being tested, in addition to a regular flash drive, is loaded as a keyboard. In theory, hackers use this behavior to implement the BadUSB attack. That is, when the flash drive is connected as a keyboard and starts typing commands that can lead to the execution of arbitrary code. This behavior of the USB drive should at least cause concern. Of course, I had a number of questions for the manufacturer.

The developers explained that this is a technical flaw, which they acknowledge and promise to fix in the next versions of the product. The fact is that during the implementation, the developers conceived a function of destruction when entering a knowingly false password. The function is very useful in many situations, especially when dealing with critical information. They require a password - you provide a knowingly false one, bam, and there is no information. And accordingly - there is no demand. The idea is great, but in our case the implementation is lame. The developers tried very hard to implement the function of feedback with the device, since neither a micro-SD drive nor a USB hub can do this, and accordingly, a controller is needed to monitor this. But what to do, because if you implement your own functionality, you will also need to install drivers, which will entail even more problems in development and raise no fewer questions about the security of the device for the end user. The guys chose the lesser of two evils and implemented an interface for interaction between the controller and cryptographic software through emulation of the HID device. I tested the device on a stand with a full USB debug and software downloaded from the manufacturer's website (as well as reassembling it myself). Despite the fact that I did not have the ability to physically read the controller firmware, I did not detect any abnormal behavior of the device. The BadUSB attack works on the principle of reproducing keystrokes. Therefore, if you simply turn off the device (at the end of the article I will write how to do this), then there can be no attack, imaginary or real, in principle. As for me, this is a technical flaw of the developers. But the choice is yours.

552a2adb-1933-4499-a8fb-ed24d4f2769b.png


Now let's look at the software. All software is written in GOlang, with all the pros and cons. The source code is published by the developer in the repository on Github. In fact, the software simply creates crypto containers on a flash drive, encrypted with Blowfish (bcrypt with a key length of 60 bytes or 480 bits, which, in principle, is quite reliable if you have a strong password). So what was all this for? After all, you can get by with regular VeraCrypt?

As strange as it may sound, both specialized software and input device emulation are implemented for one purpose only. This is the self-destruction function by a false password. How does this happen?

It would seem that if we were to implement something like this, we should store the hashes in the controller itself, thereby implementing some kind of TPM, then we could limit the number of combinations entered and implement the self-destruction function with a false password quite reliably. But the developers decided to cheat fate. From the code we see that the hash verification occurs in the software, for which the standard GO bcrypt library with a 480-bit hash is used:
Code:
storage.go:

type Hash [60]byte

func (h *Hash) Make(raw string) error {
    hash, err := bcrypt.GenerateFromPassword(
        []byte(raw), bcrypt.DefaultCost,
    )

    if err != nil {
        return err
    }

    copy(h[:], hash)
    return nil
}

It could have been more. But the main failure of the idea is not this. Let's see how the self-destruction procedure occurs when a fake password is entered:
Code:
usb.go:

func usbWrite() error {
    _, err := usbCommand("write", "0xac")
    if err != nil {
        return err
    }

    return nil
}

No checks with hardware. Passwords, both correct and obviously false, are stored in a crypto container:
Code:
main.go:

func setFakePass(opt ...*sciter.Value) *sciter.Value {
    if len(opt) != 1 {
        return Error(unkn_error)
    }

    err := storage.pass.Real.Compare(
        opt[0].String(),
    )

    if err == nil {
        return Error(pass_same)
    }

    err = storage.pass.Fake.Make(
        opt[0].String(),
    )

    if err != nil {
        return Error(unkn_error)
    }

    err = storage.Update()
    if err != nil {
        if os.IsPermission(err) {
            return Error(perm_error)
        }

        return Error(unkn_error)
    }

    return OK(true)
}

That is, everything is implemented in software. If you disable the HID device or block the command, as well as write your own software to decrypt the container (after all, the standard library is used), the self-destruction function by a false password is pointless.

Although, for some justification of the developers, it can still be noted that the self-destruction function works with the battery disconnected, provided that the device is connected to USB (how else to pick up the password? Unsolder and dump the memory chip? This also happens, in most cases not in the CIS, but if you are worried about this - then this device is simply not for you).

The case has no protection against opening. Therefore, it is better to have time to press the button. If you still did not have time to do this, there is still hope for using the function of a deliberately false password, which you can provide.

Epic fail. Hopefully it will be better in the next versions.

However, for people who will have to deal with potential attackers with low technical qualifications (criminals, tax, street police, technically unsophisticated competitors, as well as customs officers, both domestic and foreign) - a flash drive with such a function can help. Especially if you remember that even if you disconnect the battery, when entering a deliberately false password, as well as when pressing a button, the discharge will still occur (if the driver is not blocked and the original software is used) when connected to USB, this is quite enough.

Is it possible to recover information?​

Well, and the most important thing. That for which all this was started. Namely, the physical destruction of information on the carrier. How effective is it?

Let me remind you that the destruction occurs by discharging a cascade of capacitors with a peak voltage of 74 volts directly into the memory card. In my test, the voltage of the electric discharge was 68 volts.

This voltage is enough to destroy a significant portion of data. Assuming that a strong password and several containers were used for encryption, you can be sure that using standard SD card recovery tools will not lead to anything.

The laboratory I contacted also didn't help me in any way after several days of trying.

According to the developers, the first version of the flash drive was tested by SecurityLab. They also failed to restore anything.

I can say with a high degree of probability that when using several information storage containers, one nested within another, it is unlikely that you will be able to restore the information, much less brute-force anything.

What should paranoids do?​

As I wrote, the flash drive emulates an input device. And knowledgeable people understand that this is a potential springboard for a BadUSB attack. The flash drive did not show any activity, but I also could not extract the firmware from the device for analysis. This means I cannot give a 100% guarantee. However, for those who need a flash drive with a data destruction function, they can simply disable the HID device.

For Windows, PowerShell:
Code:
Disable-PnpDevice -InstanceId 'USB\VID_16C0&PID_05DF&REV_0100'

Or simply disable it in Device Management:

bb918eda-b98f-4e53-b820-f8249321af5c.png


There is nothing more to worry about.

And once again. The device was connected to the computer for several days with a complete debug of information exchange with the computer. No suspicious activity was noticed.

Verdict​

The device is usable and, if necessary, copes with its main function - data destruction. I can say with a high degree of confidence that in the CIS, it is unlikely that anyone will be able to recover information from it. What about the USA? There were cases when RAM was restored from already de-energized computers.

The flash drive manufacturers are people with many years of experience in developing information security devices, as evidenced by the article about the first version of Silenda Flash in the Hacker magazine.

If the device is technically implemented at an acceptable level, then there are serious problems at the software and logic level, for example, there is no protection against opening, and I would not place great hopes on the function of destruction by a fake password. This is a technical miscalculation of the developers, on which they spent a lot of resources. And in addition, the function was implemented very suspiciously. Nevertheless, the flash drive easily turns into a regular container with the function of deleting data. To do this, you just need to disable additional drivers.

The device works. The information from the flash drive is deleted without the possibility of data recovery even using advanced forensic tools. I can say this with a fair degree of confidence. In combination with other encryption tools, such as VeraCrypt, the device works perfectly. And this provides an additional level of data protection.

Who is this device for: for secretaries and accountants who have something to hide, for petty officials, drug dealers, dealers and even bigger dealers, as well as advanced users who just want to secure their information a little. In general, for all those who will not have to deal with technically advanced information security experts, be it cyber cops, security officers, whitehats and blackhats of any stripe. Separately, I would like to note that such a flash drive will certainly help journalists and fighters for freedom of information, given the alarming information about constant illegal searches, seizure of information carriers and pressure from law enforcement agencies in the Russian Federation and the Republic of Belarus.

Who this device is not suitable for: cyber paranoids, people who have serious secrets, a lot of information, but do not understand information security.

I have not seen absolutely protected USB flash drives yet. Want real protection? Buy yourself an HSM. Price tag from $5000. As for the device under consideration in the article: for its price (not much more expensive than a regular flash drive) 64 GB, it is quite a worthy representative of its category, perfectly coping with its intended purpose - destroying information if necessary.

  • And most importantly, remember: even the most advanced device won’t save a fool.
You can buy it on the manufacturer's website: https://silenda.systems
 
Top