replacing compromised biometric authenticators

Thu Oct 12 02:25:19 UTC 2017

I agree that multiple levels are best and, for the moment, I'd frankly
be hesitant to give anything like finger print data since one can
never change that and the harm of it getting loose can not yet be
determined. (Not that the data being taken by these scanners is
necessarily all that grandiose.)

I also would accept a facility that did something like handscan and
pin to access the lobby/security desk and keycard or fob to move
around once inside along with scan in/scan out enforcement. (No tail
gating.)

I've never really been keen on relying on biometrics though. The
handscanners can be convenient for not having to carry anything around
but when all is said and done, they are really not all that much
better than just a keycard.

-Wayne

On Wed, Oct 11, 2017 at 04:10:51PM -0500, Matt Harris wrote:
> I would definitely not say that it is current best practice not to deploy
> biometrics.  As part of a holistic approach, biometric systems can improve
> security greatly.  As a singular approach, using it as a single factor for
> authentication and authorization of access/actions, it's as terrible an
> idea as any other.  The difficult of passing a high-quality biometric
> authentication system, even knowing its success conditions, is
> non-trivial.  The good ones check for basic signs of life, as well, so
> simply cutting off someone's hand and trying to use it would fail, for
> example.  There are, of course, cheap biometric systems that are not as
> good, and ymmv depending on what and how you deploy biometrics.  Taking the
> specific threat level you're up against is always relevant.
> 
> All of the facilities I have in production have a three factor approach to
> access - "something you know, something you have, and something you are."
>  Biometrics being the latter, plus a badge or dongle, and a four digit
> code.  None of my production facilities can be access without all three.
> 
> Take care,
> Matt
> 
> 
> On Wed, Oct 11, 2017 at 4:04 PM, Ken Chase <math at sizone.org> wrote:
> 
> > (forking the thread here..)
> >
> > Biometrics are still the new hotness out in North America. Cologix whom I
> > deal
> > with in Canada has a dozen and a half odd POPs in canada/usa and I think
> > has
> > fingerprinting at all sites.
> >
> > If the current best operating practice is to avoid biometrics, why are they
> > still in use out here? Has anyone gotten the message? Is anyone in North
> > America
> > ripping them out yet?
> >
> > Other factors include your country's privacy regulations for storing
> > irreplaceable personal information, the burden of which might not be worth
> > the security 'benefit'.
> >
> > /kc
> >
> >
> > On Wed, Oct 11, 2017 at 04:46:02PM -0400, William Herrin said:
> >   >On Wed, Oct 11, 2017 at 4:32 PM, J??rg Kost <jk at ip-clear.de> wrote:
> >   >
> >   >> Do you guys still at least have biometric access control devices at
> > your
> >   >> Level3 dc? They even removed this things at our site, because there
> > is no
> >   >> budget for a successor for the failing unit. And to be consistent,
> > they
> >   >> event want to remove all biometric access devices at least across
> > Germany.
> >   >>
> >   >
> >   >Hi  J??rg,
> >   >
> >   >IMO, biometric was a gimmick in the first place and a bad idea when
> >   >carefully considered. All authenticators can be compromised. Hence, all
> >   >authenticators must be replaceable following a compromise. If one of
> > your
> >   >DCs' palm vein databases is lost, what's your plan for replacing that
> > hand?
> >   >
> >   >Regards,
> >   >Bill Herrin
> >   >
> >   >
> >   >--
> >   >William Herrin ................ herrin at dirtside.com  bill at herrin.us
> >   >Dirtside Systems ......... Web: <http://www.dirtside.com/>
> >
> > --
> > Ken Chase - math at sizone.org Guelph Canada
> >
> 
> 
> 
> -- 
> Matt Harris - Chief Security Officer
> Main: +1 855.696.3834 ext 103
> Mobile: +1 908.590.9472
> Email: matt at netfire.net

---
Wayne Bouchard
web at typo.org
Network Dude
http://www.typo.org/~web/