Malicious Cryptography - PGP

This is a discussion on Malicious Cryptography - PGP ; wandered across this site: http://www.cryptovirology.com/ it describes how malware writers can use public key cryptography towards malicious ends it does specifically mention pgp while not described as one of the examples in this site, is it possible for such attackers ...

+ Reply to Thread
Results 1 to 6 of 6

Thread: Malicious Cryptography

  1. Malicious Cryptography

    wandered across this site:
    http://www.cryptovirology.com/

    it describes how malware writers can use public key cryptography
    towards malicious ends

    it does specifically mention pgp


    while not described as one of the examples in this site,
    is it possible for such attackers to discredit the PKI by
    intentionally using people's public keys, obtained randomly off
    keyservers,
    to hide the source of their attacks ?

    tia,

    vedaal
    (am at a public workstation, so have intentionally not signed)

  2. Re: Malicious Cryptography

    vedaal@hush.com writes:

    > wandered across this site:
    > http://www.cryptovirology.com/
    >
    > it describes how malware writers can use public key cryptography
    > towards malicious ends


    It seems like a non-topic. Of course there are negative applications of
    cryptography, just as there are negative applications of every other
    technology. What's the big deal?

    > it does specifically mention pgp


    So?

    > while not described as one of the examples in this site,
    > is it possible for such attackers to discredit the PKI by
    > intentionally using people's public keys, obtained randomly off
    > keyservers,
    > to hide the source of their attacks ?


    How would anyone do that?

    --
    Transpose hotmail and mxsmanic in my e-mail address to reach me directly.

  3. Re: Malicious Cryptography

    Mxsmanic wrote in message news:...

    [...]

    > > while not described as one of the examples in this site,
    > > is it possible for such attackers to discredit the PKI by
    > > intentionally using people's public keys, obtained randomly off
    > > keyservers,
    > > to hide the source of their attacks ?

    >
    > How would anyone do that?


    the same way as they described on their site,
    but by using a public key instead of their own key

    [1] the virus includes a crypto key and symmetric key, and encrypts
    portions of the infected user's harddrive
    (this is the crypto attack described on the site, for
    crypto-extortion)

    [2] the not-too-difficult tweak,
    is for the virus-writer to 'save' the session key before it is
    encrypted to
    the public key

    [3] the virus-writer can still extort and decrypt,
    but the public key holder gets the blame/attention


    i'm not saying that this has happened,
    and really hope that it doesn't,

    am just concerned that if the malware writers start to use crypto
    together
    with their viruses/trojans,
    they may pose a bigger threat than they have until now,

    and it may be minimized if the crypto community is aware of how strong
    crypto can be misused and take/develop defensive measures


    vedaal

  4. Re: Malicious Cryptography

    vedaal@hush.com wrote:

    > [1] the virus includes a crypto key and symmetric key, and encrypts
    > portions of the infected user's harddrive
    > (this is the crypto attack described on the site, for
    > crypto-extortion)
    >
    > [2] the not-too-difficult tweak,
    > is for the virus-writer to 'save' the session key before it is
    > encrypted to
    > the public key
    >
    > [3] the virus-writer can still extort and decrypt,
    > but the public key holder gets the blame/attention


    This is basically the same attack as a malware author sending data
    unencrypted to a victim email address. The victim still has nothing
    to do with the malware, but seems guilty because it's his address
    being sent to. This attack is not changed in any meaningful way by
    the use of crypto.

    David

  5. Re: Malicious Cryptography

    vedaal@hush.com writes:

    > the same way as they described on their site,
    > but by using a public key instead of their own key


    No, that wouldn't work.

    > [3] the virus-writer can still extort and decrypt,
    > but the public key holder gets the blame/attention


    No, he doesn't. Objects are signed with the private key, not the public
    key. Encrypting or signing something with a public key doesn't attach
    any blame to anyone ... that's why it is called a PUBLIC key (because
    anyone can use it).

    > i'm not saying that this has happened,
    > and really hope that it doesn't,


    It doesn't make sense.

    > am just concerned that if the malware writers start to use crypto
    > together
    > with their viruses/trojans,
    > they may pose a bigger threat than they have until now,


    I don't see any change in threat.

    --
    Transpose hotmail and mxsmanic in my e-mail address to reach me directly.

  6. Re: Malicious Cryptography

    David Shaw wrote in message news:...

    [...]

    > > [3] the virus-writer can still extort and decrypt,
    > > but the public key holder gets the blame/attention

    >
    > This is basically the same attack as a malware author sending data
    > unencrypted to a victim email address. The victim still has nothing
    > to do with the malware, but seems guilty because it's his address
    > being sent to. This attack is not changed in any meaningful way by
    > the use of crypto.


    well,
    yes and no

    it depends on what the attacker's goals are:

    if the attacker wants to undermine the PKI system,
    and launches a large scale attack,
    and infected users find that their data can be retrieved
    by having the public key holder decrypt it for them,
    then the public key holder can be besieged enough with requests/orders
    for decryption,
    until he/she just finds it more expedient to give up the key,
    and genrate a new one that is not uploaded to a server


    my underlying intent in bringing up this topic is to ask if it is
    feasible to have a crypto analog of a 'firewall':

    a program installed by the user that alerts the user whenever a file
    is about to be encrypted,
    and asks the users permission to do so,
    (preferably with a 'hardware/smart-card' setup,
    to avoid defeat by a simple keylogger)

    the user could simply click 'yes' for all legitimate encryptions
    and prevent any malware intruder encryptions


    tia,

    vedaal

+ Reply to Thread