Language Model Generated Injection Attacks

Discussion in 'other anti-virus software' started by lordraiden, Mar 11, 2010.

Thread Status:
Not open for further replies.
  1. lordraiden

    lordraiden Registered Member

    Joined:
    Jan 30, 2006
    Posts:
    3,067
  2. gnervt

    gnervt Registered Member

    Joined:
    May 6, 2005
    Posts:
    53
    Location:
    Germany
    Many thanks for that!
     
  3. CloneRanger

    CloneRanger Registered Member

    Joined:
    Jan 4, 2006
    Posts:
    4,833
    Sounds like one for Rmus :D

    cia.gif

    Wonder how long before we see some nasties out there in the wild made using this technique ?
     
  4. Kees1958

    Kees1958 Registered Member

    Joined:
    Jul 8, 2006
    Posts:
    5,857
    It is far fetched. Problem with this theoretical approach is that program logic demands a more less fixed sequence of commands, There is off course some room for code placement, but it would be very hard to construct natural language out of it.

    Mythbusters would classify this as faintly possible, but very unlikely it has happened (will happen in practise) as the myth claims.

    This is just another (next) step in code hiding.
     
Loading...
Thread Status:
Not open for further replies.