[Owasp-topten] [Owasp-leaders] On "Insufficient Attack Protection", and the role of OWASP...

Timothy D. Morgan tim.morgan at owasp.org
Thu Apr 13 17:06:56 UTC 2017


A word of caution about "data driven":  Virtually all data in infosec is pretty
awful, from a scientific perspective.  It's plagued by subjective decisions
about up-front categorization and confirmation bias.  So your pentesting firm
gives you stats on the kinds of bugs they find?  Did they bother to look for X
class of vulns? Does your sales/marketing team encourage them to write up
findings like "you don't have a WAF" because the company also sells a WAF?
Sure, that last one sounds blatant, but this bias can be more subtle and just as
detrimental.

Ultimately, whatever data we have about what vulns are present, we don't have
the more important data to associate with it: that is which vulnerabilities are
actually being exploited?  How high of a risk are each of these flaws
typically?

You asked the parties providing data a question about what Top 10 item should be
removed... Two responded that A9 should be removed and no other suggestions
for removal were provided. Only ~12k instances of this issue was found in
the dataset. Yet this controversial category is kept, while XXE is far more
prevalent, isn't an injection attack, and is just lumped in incorrectly with
injections.

The two new categories were added, both suggested by Contrast Security and
*only* by Contrast Security.  Meanwhile, two respondents suggested adding XXE
(CWE 611) as it's own category.  The data provided makes it seem as if Contrast
Security is running the show.

To me, a better approach would be to take a broad survey of pentesters and ask
them good questions about what vulnerabilities they are seeing with more
prevalence in both legacy applications and newly written applications.  The
human mind acts as a good filter, and while no one pentester tests all types of
applications, well chosen questions and lots of responses will allow your stats
to converge to a center of mass the industry can live with.

tim


On 2017-04-13 00:21:08 +0000 Michael Coates <michael.coates at owasp.org> wrote:

> Eoin,
> 
> Very fair questions and I agree with you to look for a data driven
> approach.
> 
> I'm not part of the top 10 project so consider my feedback anecdotal at
> best.
> 
> 
> 
> 
> On Wed, Apr 12, 2017 at 4:55 PM Eoin Keary <eoin.keary at owasp.org> wrote:
> 
> > Hi Michael,
> >
> > The question for me, as a contributor to the top 10 via our SaaS/edgescan
> > is; - what metrics were used to draw the conclusion?
> >
> > Is it subjective/"whatever feels right" or is it based in data?
> >
> > Without data anyone can draft a top 10 as it means little other than a
> > marketing exercise or awareness doc not not accurate reflection.
> >
> > As an organisation we should be directing people to quick wins, most
> > common issues and focus on a risk based approach..what's the most common
> > vulnerability etc
> >
> > The issue A7 is rather unclear. Some folks are saying RASP, others WAF,
> > you are saying credential stuffing / focused brute force. So given there is
> > debate on its meaning amongst us how can we expect developers to grasp this
> > issue?
> >
> > Love you all 😍🤡
> >
> > Eoin.
> >
> >
> >
> >
> >
> >
> >
> > Sent from my iPhone
> >
> > On 12 Apr 2017, at 19:34, Michael Coates <michael.coates at owasp.org> wrote:
> >
> > There will be lots of discussion on the new Top10 RC, which is great. I
> > encourage many to bring comments, feedback and data to the conversation.
> >
> > I'll keep my comments brief. I'm very much in favor of A7. It could use
> > some word cleanup, perhaps a more fitting title too. But the spirit of
> > what's being discussed is an important advancement to defending web
> > applications. (Also, I don't consider this to be a WAF recommendation, I
> > wouldn't go that route on this at all)
> >
> > A7 reminds me of the massive credential stuffing attacks issue that has
> > hit many big sites over the past 18 months. No amount of secure dev, top
> > 10s, or WAFs stop credential stuffing. Instead you need active defense that
> > is smartly part of the application design.
> >
> > So, if A7 draws attention to this type of issue (or similar - see OWASP
> > automated threats
> > <https://www.owasp.org/index.php/OWASP_Automated_Threats_to_Web_Applications>)
> > then I consider that a win.
> >
> > Just my 2cent contribution to the larger conversation.
> >
> > Carry on security folks!
> >
> >
> >
> >
> > --
> > Michael Coates | @_mwc <https://twitter.com/intent/user?screen_name=_mwc>
> >
> >
> >
> >
> >
> >
> > On Wed, Apr 12, 2017 at 4:42 AM, Eoin Keary <eoin.keary at owasp.org> wrote:
> >
> > As a Contributing company to the Top10 stats I'd like to understand the
> > stats behind both new additions. Appreciated if someone can point me to the
> > right files/stats model?
> >
> >
> >
> >
> > Sent from my iPhone
> >
> > On 12 Apr 2017, at 05:19, Azzeddine Ramrami <azzeddine.ramrami at owasp.org>
> > wrote:
> >
> > Hi,
> >
> > I agree to change the name from "Insufficient Attack Protection" but not
> > to Improper Trust Modeling".
> >
> > I suggest to change it to "Insufficient Attack Detection and Response".
> >
> > Regards,
> > Azzeddine
> >
> > On Wed, Apr 12, 2017 at 7:24 AM, Norman Yue <norman.yue at owasp.org> wrote:
> >
> > Hey folks,
> >
> > Greetings from sunny Sydney - I hope this email finds you well. I
> > apologise for spamming owasp-leaders with this, but I think this is
> > important enough that this warrants the attention of the international
> > leadership community.
> >
> > Traditionally, we have been a trusted source of information with regards
> > to web application information security, providing both tools and technical
> > reference information to developers and application security professionals,
> > to help secure the Internet for everyone.
> >
> > Today, "Insufficient Attack Protection" is actually being considered for
> > inclusion in an OWASP Top Ten list.
> >
> > (Constructively, I think this should be replaced with something like
> > "improper trust modelling", and we push the Google BeyondCorp line of
> > thinking https://research.google.com/pubs/pub43231.html - the polar
> > opposite to "buy a waf").
> >
> > Words do not express my burning rage, and my disappointment that no-one
> > else appears to feel the same way (I read through the owasp-topten list
> > before posting this). Do people still care about the future of this
> > community, and how OWASP is perceived throughout the information security
> > industry?
> >
> > With best regards,
> >
> >
> > Norm
> >
> > _______________________________________________
> > OWASP-Leaders mailing list
> > OWASP-Leaders at lists.owasp.org
> > https://lists.owasp.org/mailman/listinfo/owasp-leaders
> >
> >
> >
> >
> > --
> > Azzeddine RAMRAMI
> > +33 6 65 48 90 04 <+33%206%2065%2048%2090%2004>.
> > OWASP CSRFGuard Project Leader
> > OWASP Leader (Morocco Chapter)
> > Cognitive Security Expert
> >
> > _______________________________________________
> > OWASP-Leaders mailing list
> > OWASP-Leaders at lists.owasp.org
> > https://lists.owasp.org/mailman/listinfo/owasp-leaders
> >
> >
> > _______________________________________________
> > OWASP-Leaders mailing list
> > OWASP-Leaders at lists.owasp.org
> > https://lists.owasp.org/mailman/listinfo/owasp-leaders
> >
> >
> > --  
> 
> --
> Michael Coates | @_mwc <https://twitter.com/intent/user?screen_name=_mwc>
> OWASP Global Board



More information about the Owasp-topten mailing list