Friday, October 25, 2019

...Drawing a Line in the Sand - Part One - Scripting Controls




MINIMAL TECHNICAL CONTROLS STANDARD PROGRAM


TOPIC = BUSINESS CRITICAL REUSABLE TEMPLATES or SCRIPTS




This document outlines the minimal technical controls for business critical reusable template files or scripts.  A reusable template [RT] is defined as a separate file that defines a set of resources. RT’s are intended to be reused across different deployments, which creates consistency across complex deployments.

CONTROLS


 ESCROW:


1.      The RT, when approved must be stored in escrow.  Priority will be given for storage in the authoritative Network Hardware Storage Module (NET HSM).

a.      If NET HSM storage is not technically feasible an enterprise secure escrow repository will be implemented with technical controls that satisfy PCI-DSS version 3.0 criteria at a minimum:

                                                              i.      Isolated Virtual Local Area Network (V-LAN).

                                                            ii.      Encryption of data at rest.

                                                           iii.      Anti-Malware protection (For example: if the host OS utilizes McAfee, the repository should implement Malware Bytes).

                                                          iv.      Physical security and access control list protections.

                                                            v.      Logging, monitoring and real time alerting.

                                                          vi.      Written standards for Create, Read, Update and Delete operations [C.R.U.D.] must be put in place to control the RT’s in escrow to ensure integrity.



RECURSIVE DATA VALIDATION:

1.      Using an automation, the RT checked out of escrow must be compared to the RT in escrow at run time.

a.      This comparison must be logged and reported on on a recurring basis.

b.      A separate account must be created to enable this automation to ensure that the audit trail supports data integrity.

2.      After modification and in accordance with currently authoritative change management processes, the contents of the RT must be verified to ensure that nothing unauthorized has been added to the RT, for example, but not limited to:

a.      Encryption Keys

b.      Identity Tokens

c.      USERID’s and/or Passwords

d.      Other security context descriptors

e.      Questionable content or sub-routines

 


AUTHENTICATION:


1.      The security context of the requestor may be evaluated at run time before allowing RT’s to be checked out of escrow (are the credentials valid or invalid?).

a.      This is applicable to human and non-human requestors.

b.      Credentials for authentication can take many forms as documented in NIST SP 800-36b, “Digital Identity Guidelines: Authentication and Lifecycle Management”, section 4.1.1.

                                                              i.      Any of these are acceptable.

c.      Periodic re-authentication may take place at a time interval that is indicative of the value or classification of the data being accessed / protected.

2.      Authentication credentials must be requested at run time from the System of Record (SOR) for authentication via the consuming system.

3.      If required, multi-factor authentication [MFA] controls will be implemented to ensure a greater expectation of trust in the identity of the requestor and ensure non-repudiation.

a.      These controls will be based on the authoritative MFA standard at the time.  

4.      Authentication events must be logged and reported on.

5.      Digital Certificates used for mutual authentication must be requested at run time and never stored in the RT, portable mass storage device or system file system(s).

a.      Certificates used for authorization must be validated at run time against the issuing Certificate Authority (CA) Certificate Revocation List (CRL).

6.      Outbound cloud provider configuration connection requests:

a.      Must verify the IP address of the requesting and requested host system and the security context of the requester (human or non-human) before completing the connection request.

                                                              i.      Ensure the IP reputation of the remote host prior to finalizing the connection.

b.      That are not encrypted will be rejected at the cloud provider perimeter or end point.

c.      Must be logged and reported on.

 


AUTHORIZATION:


1.      Authorization requests:

a.      Must be evaluated in a standard manner.

b.      Authorization logic must be protected and escrowed to ensure integrity.

c.      Must follow a standard: Policy Enforcement Point [PEP]; Policy Decision Point [PDP]; Policy Information Point [PIP] architecture.

                                                              i.      https://en.wikipedia.org/wiki/Attribute-based_access_control

                                                            ii.      https://www.oasis-open.org/committees/tc_home.php?wg_abbrev=xacml

                                                           iii.      Attribute Based Access Control:  NIST SP 800-162.



LOGGING REQUIREMENTS:


1.      Requests for AUTHENTICATION and AUTHORIZATION

a.      Successful and un-successful

b.      From the System of Record [SOR].

2.      Requests from the Escrow Repository

a.      Successful and un-successful

3.      Exception handling

a.      Successful and un-successful

b.      Calls back to sending system if validation is negative

4.      Specifically

a.      Connection requests to and from cloud vendors



GOVERNANCE:




Controls in this document are extracted from:



1.    Cloud Security Alliance, “Security Guidance for Critical Areas of Focus in Cloud Computing”, v4.0

2.    NIST, “Cloud Computing Reference Architecture”

3.    Cloud Security Alliance, Controls Matrix

4.    NIST, Zero Trust Architecture

5.    NIST, Cyber Security Framework

6.    AXELOS, IT Service Management Framework, v3  

RACI Matrix





Minimal technical Controls for Cloud Deployment Scripts




[1] https://cloudsecurityalliance.org/artifacts/security-guidance-v4/

[2] https://www.nist.gov/publications/nist-cloud-computing-reference-architecture

[3] https://cloudsecurityalliance.org/research/working-groups/cloud-controls-matrix

[4] https://csrc.nist.gov/publications/detail/sp/800-207/draft

[5] https://www.nist.gov/cyberframework

Tuesday, October 15, 2019

...please don't ask questions that you don't want an honest answer for...




Artificial Intelligence or AI?  Not ready for prime time...will it ever be? 

It would take a Phil Zimmermann, PgP type, "black swan" event to make it a reality. 

"...True A.I. is not logically impossible, but it is utterly implausible....".

What possible financial incentive would there be for a for profit business to develop a product, that when sold, would by design, improve itself, constantly, never needing a maintenance contract, or a patch? 

None.

The article below, not the best analysis, but, certainly a reasonable, rational, thought provoking analysis.


#artificialintelligence #hypecycle
 

Copyright © 2019 by"the Secret CISO"

All Rights Reserved.

 

Thursday, September 26, 2019

...Gone in 60 Seconds - redux...




Reality is a harsh teacher. 

Taking the easy way out, "Check the Box" security carries with it a terrible cost. 

The link below outlines in non-technical terms, how reconnaissance and attack via ransomware took the world’s largest shipping company, Maersk, down crippling their worldwide operations in less than sixty seconds (the actual attack that is, the reconnaissance phase?, must have been substantial). 

80,000 employees, in 574 offices in 130 countries across the globe, seventy-six major ports and 800 ships at sea, DEAD IN THE WATER...

They appear to have been big fans of “check the box” security, I wonder how they feel about those decisions now?

The untold story of NOTPETYA malware and MAERSK shipping

Copyright © 2019 by"the Secret CISO"

All Rights Reserved.

 

Monday, September 23, 2019

...Nation State throws down the gauntlet on "offensive cyber operations"...

French armed services minister Florence Parly, who unveiled France's new offensive cyber doctrine


France, a permanent member of the UN Security Council and a nuclear power, releases it's national policy on "...Offensive Cyber Operations..." relevant to peacetime and open hostilities. 
 
Amazing reading. 
 
Seldom to Nation States make such statements. 
 
Take the time to translate and read, please post a comment if you find an accurate translation. 
 
 

Copyright © 2019 by"the Secret CISO"

All Rights Reserved.

 

... an API security standard, about damn time!...

Above is my humble entrant for the API Security Top Ten challenge


F I N A L L Y  we have a "top Ten" List for API security, the OWASP list is almost as good as my home grown list...almost.

API Security Reference Architecture - OWASP finally publishes API Security "Top Ten" list. 
 
Reminds me of Neil Armstrong stepping on the Moon for the first time.
 
 

Copyright © 2019 by"the Secret CISO"

All Rights Reserved.

 
 
 

Wednesday, August 14, 2019

Defending against Ransomware? That's easy, try eating soup with a knife, now, that's hard!


Why all of the fuss?

 

Ransomware it's on the uptick in the Gartner "hype cycle". People want to sell you lots of shiny new tools that you must have to defend yourself? Why would I go to the CFO and recommend that? People will swear to you that security is "better in the cloud". It is not, they are just attempting to sell you something. Resist the temptation.

Success with ransomware defense is not a function of your tools, your network topology, or your SIEM. A successful defense against ransomware is determined by your ability to make rational decisions and move left of bang thru the "political will" minefield.

OK, now you can send me flame mail.

 

In his truly seminal work, "The Laws of Human Nature", Robert Green walks us through the biases that will turn your decision making away from rational thinking and enable you and your team to embrace your biases: Confirmation Bias; Conviction Bias; Appearance Bias; Group Bias; Blame Bias and Superiority Bias.

The criminals that will use ransomware to steal from you will use your normalcy bias against you. Don't let that happen. Look your lack of political will squarely in the face and say, "...not today...".

To effectively mitigate ransomware risk, we must cultivate the political will, right now, to take reasonable, cost effective action. Remember these wise words: "Cyber Security is Everyone's Job".

M&M candy or a box of rocks? You decide.

 

Most of us have been trained that "defense in depth" is the be all - end all of Information Security. I call that "the M&M candy" methodology: have a hard crunchy shell to protect the soft tasty center.

Most of the businesses I work with allow non-security personnel to make security decisions because these decision makers do not understand security and risk and refuse to trust the security professionals they hire. This is cognitive dissonance on a massive scale.

What I am proposing here is a paradigm shift [really just an over simplified outline] to what I call the "box of rocks" methodology: everything in the box is as hard as a rock and you just can't take a bite without breaking your teeth.

How can you become a box of rocks?

 

Below is my, "box of rocks" approach to ransomware defense. Granted it is not all inclusive. It is also not expensive. You probably have all of these capabilities already. If you do, you are most likely not using the capabilities below in an all-inclusive manner.

If your IT folks tell you that this approach is crazy - hire better in the future.

 



As always, if you follow the outline below and you get a successful ransomware attack prosecuted at your customer, that's a "you thing".  If you want to ask questions about any of the ideas below, just send us an email via the comments section.

What we need to get this party started

 

A threat intelligence data feed [TIDF]

 

We'll need a threat intelligence data feed that incorporates near real time DNS and IP address white/black listing that can be consumed by all of your perimeter defense systems and your internal routers.

This allows you to mitigate the risks of vendor provided TIDF's that are universally mediocre, normalize protections across your perimeter and internal traffic management systems.

Do not default to accepting your firewall vendors schtick that their TIDF feed is awesome!

This is very high payoff and very cost effective.

A configuration management database [CMDB] capability with teeth

 

You'll need a CMDB that is integrated with your change management system, your vulnerability management system, your patch management system (and their supporting processes) with enough intelligence to monitor your key systems and autonomously roll back unauthorized configuration changes, in real time, without human intervention.

I'll bet this is a capability of your Service Desk Ticketing software package?

Building this is worth your time. .

Security Orchestration, Automation and Response [SOAR]

 

Though this is a major player now in the hype cycle, this actually drives incredible business value via risk mitigation. If you have a SIEM solution in place you probably don't have to purchase anything, chances are good that you already own the parts and pieces, you just need to do some plumbing.

Imagine a time when your employees bring in an infected device, plug it into your network, your internal router, consuming your TIDF sees the outbound connection request to a threat actor, and autonomously, without human intervention, in nanoseconds, isolates the threat from your business and dispatches a human to investigate.

Think of it like "Sky Net" but, cuddly and smelling like fresh baked chocolate chip cookies.

Integrate your SOAR tool kit with your network vendors "Network Access Control" capabilities, and your Security Information and Event Management capabilities.

Building this is worth your time.

An easy to use Password Manager that employees can't live without

 

We all talk a good game about password management but do any of us have an enterprise wide password manager that your employees love?

Nope.

One of my larger, international customers was also skeptical of this, but, once it was in place, it was fervently embraced by the employees simply because they could use it for personal use. Being a centrally managed, enterprise platform that synch'd desktop, phone, etc...passwords. We could mandate twenty-one character complex passwords. Automatic password changes, none of the employees had to lift a finger, they L O V E D it!

No more forgotten password Service Desk Telephone Calls!

All accounts, including service and daemon accounts are now, at a minimum, twenty-one characters and complex.

A major step in the right direction.

Inexpensive, a major risk mitigation.

Protection Profiles for high risk groups of employees

 

Check box / one size fits all security solutions waste resources (labor and money). Develop "Protection Profiles" - bundles of minimal technical controls, tools, capabilities, for groups of high risk employees.  They get to feel like James Bond, they will love it.

Make Multi-Factor Authentication a key component of your first "political" Protection Profile group, their adoption of these James Bond style "cool kids" capabilities will make it political kryptonite for nay-sayers re: future adoption.

Easy to explain to the Board, CEO and CFO. It just, plain and simply, makes great business sense.



Define and Implement Minimal Technical Controls for your public facing web servers and API's

 

Incorporate the OWASP top ten controls into your public facing web presence. Introduce in phases, getting closer to our mythical "box of rocks" each quarter.

Building this is worth your time.

Define and Implement Minimal Technical Controls for your internal and external web services, capabilities and protocols

 

Develop your own, "Secure Cloud Computing Architecture [SCCA]" plan, mature it each quarter.

Start out slow and pick up speed as everyone becomes comfortable with the capabilities.



Privileged Access Workstation [PAW] Model

 

Adopt a PAW approach, use it, mature it. Never administer a zone of "high trust" from a zone of "low trust".

Building this is worth your time.

Airgap your crown jewels

 

Basing disaster recovery and business resumption planning [DR/BRP] around traditional VMWare solutions in a remote data-center makes the assumption that your attackers are stupid. They are not, they are highly skilled criminals that run highly profitable businesses. They will break into your backup system, follow the bread crumbs to your DR/BRP site, encrypt your backups, then attack your online, operational systems. Pants around your ankles, you will be gasping for breath, or....you can take proactive steps, NOW.

You need to thoroughly document your key business processes, understand how they are "plumbed" thru your data center and cloud environments.

What will it take to restore these business critical, business processes in a disaster?

Who cares about your data backups if you don't really know how the business process is plumbed and you cannot re-create this plumbing to make your business processes function?

Think about it...

The world has evolved beyond servers in your data center or cloud. It's all about understanding and protecting your business critical - business processes.

Don't believe me? Call your CFO and ask him/her...

Building this is worth your time.

Implement Microsoft's "Red Forest" approach to protect Active Directory

 

Imagine what it will be like the morning NO ONE can log in...to....anything...don't be that person.

Implement "Zones of Trust" within your enterprise

 

Using the FIPS 199 approach, identify what is truly important, segment the truly important stuff into "zones of trust", utilize your labor dollars and budget to protect what it truly important. Ignore the rest until you have some unstructured free time...(yes, that is a joke)

Develop a program to manage - Ports, Protocols and Services - Normalizing, Documenting and Monitoring Network Traffic between zones of trust

 

Review the Defense Information Security Agency [DISA] "ports, protocols and services" methodologies for reference.

Integrate with your SOAR Program.

Effective Messaging Management Practices

 

Your IT folks need to understand how the river that is your network, flows thru your enterprise and where to focus your controls to minimize risk.

Do your email "dispatch" process rules, effectively scan email or chat as it moves from desk to desk?

Do your email "external" processes effectively implement "mail transport rules" that scan email inbound and outbound from your business?

If you asked your IT guys these questions what would they say?

Do you have the political will to convert all inbound embedded URL's into simple strings of text, making them "un-clickable"? To watch your phishing concerns become a thing of the past?

Empower your Security Department to take control of Security "stuff"

 

Fund and staff a "Security Architecture" team that is autonomous and reports to the CISO or CIO.

Listen to them.

Develop an actively managed, structured, system and application software hardening program. Use the CIS baseline library to increase hardening on a quarterly cycle.

Implement an ITIL RACI matrix approach, making your Security Department "accountable" and your IT department "responsible".

Your Board of Directors will thank you.

Implement Centralized Authentication and Authorization [AuthN & AuthZ] for your UNIX systems

 

Integrate this into your Active Directory infrastructure.

Implement Centralized Authentication and Authorization [AuthN & AuthZ] for your NETWORK systems

 

Integrate this into your Active Directory infrastructure.

Implement RFC 3580 Network Access Control [802.1X] - Integrate it with your SOAR program

 


It only takes six minutes for the light to move from the Sun to the Earth - when it comes, your ransomware spicyness will hit faster.  Don't buy your sunscreen after you have skin cancer.

 

Ransomware and other forms of "denial of service" attacks are highly effective, highly automated forms of attack.

Stop hoping for the best.

Plan for the worst - now.



Best of luck to all of you.

We're going to need all the luck we can get, but, what we need much more, is lots of hard work and sweat, now.

Thanks for reading, you can send me flame mail now.

Copyright © 2019 by"the Secret CISO"

All Rights Reserved.