[mdlug] [Fwd: [opensuse-offtopic] And now the Manchurian microchip]

Tue Feb 3 17:40:01 EST 2009

Michael Rudas wrote:
> --- Aaron Kulkis  wrote:
> 
>> This is as startling as in the late 80's when George Goble
>> (author of the first multi-CPU Unix kernel, and builder of
>> the first dual-VAX to run it on) came back from a security
>> conference visibly distressed. I asked him what was up, and
>> he said that he learned that he could sit at his computer
>> in a windowless room with the door locked, and someone
>> across the hall, or even across the street, could read
>> his (or anyone else's) computer screen in real time.
> 
>> [I *BELIEVE* that method won't work with LCD screens -- it
>> relies on detecting the electric field made by the electron
>> guns, and reconstructing the display]
> [snip]
>> And now the Manchurian microchip
> 
> The first one IS plausible, though not as much a risk as you might
> think.  As you correctly surmised, LCDs are invulnerable to this sort
> of hack--but even a CRT's "leakage" would max out at about 20 feet (if
> that much).  As a former TV & computer-monitor repairman, I can state
> this with a very high degree of confidence.
> 
> But I call "Bullshit!" on the second one.  First, few chips are
> actually DESIGNED in China, though several silicon foundries are
> located there--most designs come from the US, Taiwan, Europe, or
> Japan--I sincerely doubt that the PRC government could "sneak in" the
> circuitry to create such a back door, let alone integrate it AND hide
> the OS-agnostic code to make it active without it being discovered
> early on.
> 

Reply I received from a field-grade officer currently stationed in S. Korea:

-------- Original Message --------
Subject: RE: [Fwd: [opensuse-offtopic] And now the Manchurian microchip]
Date: Sun, 1 Feb 2009 08:52:16 -0700
From: Charming Rogue <rake_60 at hotmail.com>
To: 'Jeff Buhrt' <buhrt at aftinc.net>
CC: 'Aaron Kulkis' <akulkis00 at gmail.com>
References: <4984DF72.9070006 at gmail.com> <498523AC.40302 at aftinc.net>

Jeff,

The problem is real and huge.

Those in the know...know.  Lot's of uninformed speculation, whining by DoD
people who don't understand what's happening, etc.  

My personal guess is that the Chinese are doing something that we did
first...such is the problem of relying so heavily on technology.

BTW, I am an acquaintance of Aaron's...but please don't 'reply all' (or
remove me from future emails)

Aaron, please send your emails out 'BCC' or such so that my email is masked.

My best,
Mark

-----Original Message-----
From: Jeff Buhrt [mailto:buhrt at aftinc.net] 
Sent: Saturday, January 31, 2009 9:23 PM
To: Aaron Kulkis
Cc: mdlug at mdlug.org; Ryan Bansemer; Fisk, Wesley L CPL NGMI; Richard Stoll;
PATRICIA STOLL; sharon kulkis; Darryl Kulkis; Mike McDaniels; dusty bidwell;
andrew stoll; Brian Manley; Charming Rogue; Decoster, Vaughn A CPT RES USAR
USARC; Donald; baochen at gtn.ru; Frank Economou; Frank Economou; Gerard Kehoe;
Fred Hershberger; Julie Sasaki-Hom; Kirley Peter J SFC MIARNG; Kramer, Kurt
S SSG RES USAR AMC; Lewis Campbell; MAJ Wagh; MAJ Raymond Stemitz; Marco
Cuenca; Mark Cunningham; Matt Churchill; robert.mooresr1 at us.army.mil; SFC
Timothy Artibee; Thomas N Mouat; Todd Randall; Tom Vizzini; Tracy Kehoe;
Valentin Nemkov; Vernard Foley; Weems, Jeffrey T SGT RES USAR USARC; eljwc;
mike fortuna; MSG Jon Sawyer; MSG Whittaker; OURFUNNYFARM at peoplepc.net;
r.w.rynca at us.army.mil; ProtestWarrior; SSG Jack Vliet
Subject: Re: [Fwd: [opensuse-offtopic] And now the Manchurian microchip]

I have heard references to the concept of rogue chips. The question is
how to trip them and more-so how to make them communicate out without
the OS knowing... As for technology theft, think of all the graduate
students, 'cheap' H1-B's, etc. that are in US companies. I heard that 
they send 1000's of people who collect little bits of knowledge, then
back in China uses large amounts of human labor to collate intel.
The real point of attack/weakness is most of our stuff is made in China. 
I still think firmware vs chips are more at risk, not that a false path 
couldn't be embedded in a pre-loaded firmware on a board.

One of my contract engineers (and friend), has higher end equipment made
with only test software and a USB port. When it comes to the US, only
then load it before shipping or send software to the customer for them
to flash with the newest (functional) software. China or other country
of origin doesn't have the device firmware. This helps two ways: one is
security, the other reduces stolen/knockoff devices. We have also been
doing more work on embedded systems using a licensing product we
developed for a customer. [For example a large Fortune 50 customer found
their embedded link track controller firmware on ebay, which was
promptly requested to be pulled.] We also had a discussion of some new
medical equipment that is subject to 'remote' attack from wifi/Bluetooth
range.

This will be a good topic at InfraGard this week and for Wednesday's
embedded engineering group lunch here in Indy.

As a related side note, let me know if you do SOHO networking, we are
working on commercializing our secure, network load balancing
VPN/bridging appliance. We initially designed it to secure bridge
medical equipment that has non-routable network [broadcast] traffic.
Our idea is that it would work well to secure a person's at home office
to secure traffic back to their office and at the same time allow
bonding a DSL+cable connection for speed and redundancy, just plug their
PC(s) and SIP phone(s) in.

Thanks Aaron for the reference.

-Jeff

Jeff Buhrt
Achievement Focused Technology, Inc.
http://www.aftinc.net
buhrt at aftinc.net
317-843-4444 ext 107 Office
317-513-3238 Cell