[LINK] The Degree of Peril in an Insecure Wifi Network

[Paul Brooks]

On 18/07/2011 1:16 PM, Roger Clarke wrote:
> Good points, thanks Paul.
>
> But:
>
> (1)  I have my firewalls configured on the individual devices within 
> the subnet,
>       which reduces the risk.  But, agreed, this means more is visible 
> - and I guess
>       maybe a lot of people *don't* configure firewall functionality 
> on their devices)
1) Most computer-style devices have variable levels of 'firewall' function that can be
selected - the 'home' level provides significantly more access than the 'Internet
cafe' level.
Most people, given the option, will (quite reasonably IMO) select 'home' for a
connection that is inside their home,  so that their photo sharing, unified music
libraries etc all become seamlessly visible to other devices inside the home. Very few
people will set their firewall settings to the most paranoid while inside their home
network, because it breaks the very functionality they like to achieve - to make the
network invisible and open to their other devices.

2) Show me the firewall configurable functions on your IP-enabled blood glucose meter,
or the smart whiteboards appearing in classrooms over the past few years, or your
Playstation/XBox/Wii and DVD player. Not every device has inbuilt firewall functions,
and in the future M2M / Internet-of-Things it will become less common, not more. The
proverbial IP-enabled light switches don't have much of a UI for setting firewall levels.
 

> (2)  this attack pattern doesn't scale.  It's suitable for anonymous 
> access (although
>       maybe not if you're using your next-door neighbour's Wifi!), and 
> for planting
>       evidence with the intention of embarrassing and inconveniencing 
> an opponent.
>       But if you want lots of service or you want to attack lots of 
> targets, you'd
>       be better off operating over the Internet rather than doing 
> everything from
>       the (dis)comfort of your car

Sure - but in the context of the original article, and your question on comments from
Mark Gregory about the risks of open WiFi networks, you asked about relative ease of
breaking in through the two vectors. If the intention is to leach off your connection
to the Internet and download questionable content so that it can't be traced to their
device (which was the thrust of the original article), then the assumption surely is
that the offender is already parked at the bottom of your driveway - scale is immaterial.

In any case, I'm not about to open up my WiFi on the basis that a rational attacker
won't try to access my network and compromise my devices because it doesn't scale as
well as other attack methods!


> On 16/07/2011 11:16 AM, Roger Clarke wrote:
>>>  ''All of the detrimental effects of being hacked will then follow,
>>>  except the hacker has been given an easy and exploitable way into the
>>>  network,'' Dr Gregory said.
>>>  [Isn't Gregory confusing rather separate things here?  Is it
>>>  significantly easier to break into a device via a wifi network than
>>>  over the Internet connection?  And even if it is, does that approach
>>>  scale sufficiently to make it worth a miscreant's while using this
>>>  approach rather than mounting the attack over the Internet?]
>