[cabfpub] Voting Begins: Ballot 221: Two-Factor Authentication and Password Improvements

Mike Reilly (GRC) Mike.Reilly at microsoft.com
Thu May 24 19:34:43 UTC 2018


Microsoft votes NO on Ballot 221 given the procedural issues discussed regarding lack of a redline, which would make this challenging to enforce as a valid ballot.  We also have concerns related to implications of the wording on the password requirements.  Specifically:


  *   How would auditors verify and prove that a CA did not change a password more frequently than two years? This is trying to prove a negative.
  *   What about when a CA employee leaves who knows the password which requires it to be change in less than two years?
  *   What about if the password is compromised and needs to be changed in less than two years?

I did talk with Tim on the phone yesterday and apologized for not chiming in earlier with these concerns.  My initial read of the proposed ballot didn't raise these concerns and as I shared the ballot language more broadly within Microsoft and with auditors these discussion items came up.

I do appreciate Tim's and the Network Security WG's efforts on behalf of the CABF and again apologize for chiming in during voting rather than the discussion period.  I'm happy to work on improved language with the ballot drafters.

Thanks, Mike

From: Tim Hollebeek <tim.hollebeek at digicert.com>
Sent: Wednesday, May 23, 2018 5:34 AM
To: Mike Reilly (GRC) <Mike.Reilly at microsoft.com>; CA/Browser Forum Public Discussion List <public at cabforum.org>
Subject: RE: Voting Begins: Ballot 221: Two-Factor Authentication and Password Improvements

Mike, it's a SHOULD, so we have two years to fix it.  I agree the wording is awkward, it was hard to write.  I found words like minimum and maximum to be very confusing as they would get misread as the other.  There was discussion about this exact language, including at the face to face.  Your description agrees with the intent, and is what I explained at the face to face in Virginia.  Use great passwords, and DON'T overly frequently rotate them.  So you guys are correctly understanding the requirement.

On the policies for zones, this is consistent with the guidance in NIST SP800-63 that if all other things are equal, longer passwords are stronger.  So in zones with higher security requirements, we have higher length requirements.

Voting NO would prevent removal of the language that requires frequent password changes.  Which means that bad practice continues for a bit longer.

The discussion period for this ballot was quite long.  I would encourage people to not wait until the final draft and/or voting to analyze draft ballots.  I would prefer not to tweak this ballot and revote it, but I'll do that if needed.

I would have been willing to provide a redline to assist in voting if the request had come in earlier, like not a day or two before voting ends.

There was an earlier incident this year where I had to correct the redline; I'm currently erring on not including a redline if I'm not 100% (-epsilon) sure that it agrees with the ballot text.

Some of you that have been around for a while remember that redlines used to be uncommon.  The fact that they're common right now is a side effect of the fact that I write most of the ballots.

Hope some or all of this helps.

-Tim

From: Mike Reilly (GRC) [mailto:Mike.Reilly at microsoft.com]
Sent: Tuesday, May 22, 2018 10:06 PM
To: Tim Hollebeek <tim.hollebeek at digicert.com<mailto:tim.hollebeek at digicert.com>>; CA/Browser Forum Public Discussion List <public at cabforum.org<mailto:public at cabforum.org>>
Subject: RE: Voting Begins: Ballot 221: Two-Factor Authentication and Password Improvements

I'm having internal conversations with concerns regarding section iv which states:

" iv.   If passwords are required to be changed periodically, that period SHOULD be at least two years.  Effective April 1, 2020, if passwords are required to be changed periodically, that period SHALL be at least two years."

Sorry I didn't catch this during the discussion period but the language should be adjusted. A period of at least two years means you wait a minimum of two years. Can we use words like maximum?

The reference to NIST SP 800-63 makes us think we want to emphasize stronger passwords while minimizing the frequency with which those secrets are changed. That is what the NIST guidance calls for. The language "if passwords are required to be changed..." also makes us think this. Basically, password changes are optional, but if you require them to be changed, two years is the minimum amount of time to wait before changing them.

Also, why have separate policies for different security zones? A 12-character password of seahawks2018 shouldn't be considered better than an 8 digit randomly generated one.

I agree with Wayne that a redlined version should be provided for a vote as it's difficult to clearly see the changes.  I'm inclined to vote No at this point and apologize for not looking harder at this in the discussion period.

Thanks, Mike

From: Public <public-bounces at cabforum.org<mailto:public-bounces at cabforum.org>> On Behalf Of Tim Hollebeek via Public
Sent: Thursday, May 17, 2018 2:48 PM
To: CA/Browser Forum Public Discussion List <public at cabforum.org<mailto:public at cabforum.org>>
Subject: [cabfpub] Voting Begins: Ballot 221: Two-Factor Authentication and Password Improvements


Ballot 221: Two-Factor Authentication and Password Improvements

Purpose of Ballot: The Network Security Working Group met a number of times to
improve the Network Security Guidelines requirements around authentication,
specifically by requiring two-factor authentication, and improving the password
requirements in line with more recent NIST guidelines.

While CAs are encouraged to improve their password requirements as soon as
possible, a two year grace period is being given to allow organizations to
develop and implement policies to implement the improved requirements, especially
since some organizations may have to simultaneously comply with other
compliance frameworks that have not been updated yet and are based on older NIST
guidance about passwords.

The following motion has been proposed by Tim Hollebeek of DigiCert and endorsed
by Dimitris Zacharopoulos of Harica and Neil Dunbar of TrustCor.

- MOTION BEGINS -

This ballot modifies the "Network and Certificate System Security Requirements"
as follows, based upon Version 1.1:

In the definitions, add a definition for Multi-Factor Authentication:

"Multi-Factor Authentication: An authentication mechanism consisting of two or
more of the following independent categories of credentials (i.e. factors) to
verify the user's identity for a login or other transaction: something you know
(knowledge factor), something you have (possession factor), and something you
are (inherence factor).  Each factor must be independent.  Certificate-based
authentication can be used as part of Multifactor Authentication only if the
private key is stored in a Secure Key Storage Device."

Capitalize all instances of the defined term "Multi-Factor Authentication".

Add a definition for Secure Key Storage Device:

"Secure Key Storage Device: A device certified as meeting at least FIPS 140-2
level 2 overall, level 3 physical, or Common Criteria (EAL 4+)."

In section 1.j., capitalize Multi-Factor Authentication, and strike the
parenthetical reference to subsection 2.n.(ii).

In section 2.f., add "(for accountability purposes, group accounts or shared
role credentials SHALL NOT be used)" after "authenticate to Certificate Systems".

Change section 2.g. to read:

"g. If an authentication control used by a Trusted Role is a username and password,
    then, where technically feasible, implement the following controls:
  i.           For accounts that are accessible only within Secure Zones or High Security
               Zones, require that passwords have at least twelve (12) characters;
  ii.          For authentications which cross a zone boundary into a Secure Zone or High
               Security Zone, require Multi-Factor Authentication.  For accounts accessible
               from outside a Secure Zone or High Security Zone require passwords that have
               at least eight (8) characters and are not be one of the user's previous
               four (4) passwords; and implement account lockout for failed access attempts
               in accordance with subsection k;
  iii.        When developing password policies, CAs SHOULD take into account the password
               guidance in NIST 800-63B Appendix A.
  iv.         If passwords are required to be changed periodically, that period SHOULD be
               at least two years.  Effective April 1, 2020, if passwords are required to
               be changed periodically, that period SHALL be at least two years."

In section 2.h., change "Require" to "Have a policy that requires"

In section 2.i., change "Configure" to "Have a procedure to configure"

Change section 2.k. to read:

"k. Lockout account access to Certificate Systems after no more than five (5) failed
    access attempts, provided that this security measure:
  i.           is supported by the Certificate System,
  ii.          Cannot be leveraged for a denial of service attack, and
  iii.        does not weaken the security of this authentication control;"

Change section 2.n. to read:

"Enforce Multi-Factor Authentication for all Trusted Role accounts on Certificate
Systems (including those approving the issuance of a Certificate, which equally
applies to Delegated Third Parties) that are accessible from outside a Secure Zone
or High Security Zone; and"

- MOTION ENDS -

The procedure for approval of this ballot is as follows:

Discussion (7+ days)

Start Time: 2018-03-28  15:00:00 EDT

End Time: 2018-05-17 17:45:00 EDT

Vote for approval (7 days)

Start Time: 2018-05-17 17:45:00 EDT

End Time: 2018-05-24 17:45:00 EDT

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cabforum.org/pipermail/public/attachments/20180524/49e38c3d/attachment-0003.html>


More information about the Public mailing list