SlideShare a Scribd company logo
Digital forensics and giving evidence
With Jonathan Haddock
Show and tell -
digital forensics and
giving evidence
Jonathan Haddock
March 2020
About me
● I’ve worked in IT for over 15 years
● Now the Deputy Senior Information Security
Officer for three local authorities
● Have run penetration testing courses with the
YPISG
● Trained in forensics, penetration testing and
incident response
https://keybase.io/joncojonathan
@joncojonathan
https://uk.linkedin.com/in/jonathanhaddock
What is digital forensics?
● Process of looking at computer systems and
files to determine what activity took place
○ Can also be about recovery
● Increasingly important for incident response
● Not like it's shown in the movies!
○ You can't decrypt part of a drive like that
● Isn't “wet” forensics - we're in the digital realm
● Needs translation and explanation
Why translation?
● Logs can be really hard to read (or just plain
horrible!)
This just means someone accessed a file on a site, and
got redirected (resource permanently moved)...
188.65.57.4 - - [09/Oct/2018:20:32:40 +0000] "GET
/RSS/blogfeed.xml?blogID=1 HTTP/1.1" 301 185
"http://example.org/RSS/blogfeed.xml?blogID=1"
"SimplePie/1.2 (Feed Parser; http://simplepie.org; Allow
like Gecko) Build/20090627192103"
What digital forensics isn't
● Not an instant indicator of blame
○ Also not your fault
● (Often) Unable to conclusively tie actions to a
human being
○ Unless there's a camera pointing at the user
and the screen, you can't know for certain
who did what
● Not as easy as it used to be thanks to encryption
Forensic process
● Obtain the assets
○ Files, disk images or physical media, phones
● Calculate the asset’s hash
● Take an image of the asset
● Calculate hashes
● Make a copy
● Check its hash
● Perform analysis
Chain of custody
● First entry:
Assets received by you from provider
● Shows who had access, so possibility of
tampering reduced
● If the asset moves, is copied or shown to
someone else, note in the log
● Protects you
Taking hashes
● A hash is a mathematically calculated
signature for a file
○ da39a3ee5e6b4b0d3255bfef95601890afd80709
● md5, sha1, sha2
● Note tool version in your report
○ sha1sum (GNU coreutils) 8.29
● The sum will change as soon as the file
contents is edited
Taking an image
● Not just a photograph!
● A block by block copy of a disk
○ Tools like dd or ftkimager
○ Not a copy of the files on the disk
● Means we can leave the original untouched
● Beware of Windows - it’s not forensically safe
○ Use a write blocker
Copy vs
Data carve
Make a copy of the image
● Immediately after the original is received or the
disk image created
○ Making disk images takes time, you don’t
want to do it again
● Check the hash
● Move the original to protected storage
● We’ll work on a copy
● You may have to provide a copy
Forensic tools
● Tools like Encase or Forensic Tool Kit (FTK) are
expensive - around $3,000 each
● The Sleuth Kit is open source
○ Has a graphical front end - Autopsy
● PhotoRec (free) can recover files
● But sometimes you just have to use your
everyday applications
Analysing the evidence
● Pick the relevant tool
○ Cost
○ Task
● Stick to the specification
○ If looking for files or events in one date range,
don’t get distracted by others
● Try to avoid changing the copy - avoids
confusion
● Take notes / keep a journal
Writing reports and statements
● Take information from your journal / notes
● Write for the audience - probably not too
technical
○ Internal or external?
● Appendix:
○ List of hashes
○ Maybe chain of custody
● Also need-to-know
Why “need to know”?
● We have a duty of care to everyone involved in
the case
○ Prosecution
○ The accused
● Leaked details can still ruin someone’s life
○ Newspapers often publish details before
court judgments
● UK law: innocent until proven guilty
Court
Attending court
● You’re not allowed in before giving your evidence
● Expect to sit around a lot
○ Courtroom might be triple booked
● Look at your questioner
● Answer to the Magistrates
○ Watch their pens - don’t talk too fast
● Make sure your phone is off!
Be clear, explain
● Magistrates are lay people
● Your job is to explain what the evidence means
● You might be allowed a copy of your statement
○ Can’t be annotated
● If you don’t understand the question, get it
repeated or rephrased
Don’t make stuff up!
● You either found it in the evidence, or you didn’t
● If you don’t know, say you don’t know
○ If you can’t remember, say that too
● Answers should clarify or explain your
statement
○ You’re not creating new information, just
explaining what’s already presented
Cross examination
● The opposition barrister may want to ask
questions
● Their aim is to help their client, don’t feel attacked
○ Consider their questions as just other questions
● You may be asked for clarification to your earlier
answers
● Alternatively, there may be entirely new questions
● The first barrister may ask additional questions
Assumption trap
● “Is it fair to assume that…”
● Beware - is it fair to assume anything in criminal
proceedings?
● Ultimately we’re not there to speculate
● Can be a way to use your expertise to cause you to
lend support to a tenuous idea
Reporters
● There’s likely to be reporters somewhere
○ Courtroom (public gallery)
○ Waiting area (public space)
○ Just outside
○ May be in disguise
● Don’t be afraid to say “no comment”
● They can be pushy
Any questions?
Thank you for
coming
Interested in security? Try a podcast
Smashing Security
Humorous run down of security news
for the week. With Graham Cluley
(@gcluley) and Carole Theriault
(@caroletheriault).
www.smashingsecurity.com
Security Now
A semi-technical dive into recent
security news and vulnerabilities.
Hosted by Leo Laporte and Steve
Gibson (@sggrc).
www.twit.tv/sn
The slides that follow were not presented at codeHarbour due to
time, but give you the rest of the story.
I also blogged about this at
https://blog.jonsdocs.org.uk/2018/11/12/show-and-tell-digital-
forensics-and-giving-evidence/
Hash example
● Simple text file, says “hello”
f572d396fae9206628714fb2ce00f72e94f2258f
● Text file edited to say “hell0”
40dfb682064a88af16ddb3faa2d3b5f0e6a82ee3
● A minor change has completely changed the
hash
Validate your tools
● As tools receive updates their output can
change
● Have a known sample
● After an upgrade, confirm the tool still gives
known results
○ If file spoons.jpg was found before, it still
should be, and be identical
● New versions may add new features, more
results
Keeping track
● I always keep a journal
○ Details of assignment
○ Findings (each artefact has a hash)
○ Screenshots
○ Thoughts
○ Chain of custody
● Journal should be shared on a need-to-know
basis
References and acknowledgements
● Screenshots of the San Bernardino case reporting:
○ https://www.theguardian.com/technology/2016/feb/22/tim-cook-apple-
refusal-unlock-iphone-fbi-civil-liberties
○ https://www.bbc.co.uk/news/technology-35593048
● USB write blocker image: https://www.cru-
inc.com/products/wiebetech/usb_writeblocker/
● Autopsy screenshot from
https://www.sleuthkit.org/autopsy/images/v3/overview.png
● Podcast logos © their respective owners, from their respective sites

More Related Content

Similar to Digital forensics and giving evidence by Jonathan Haddock

Computer Security For Activists & Everyone (Oct 2018)
Computer Security For Activists & Everyone (Oct 2018)Computer Security For Activists & Everyone (Oct 2018)
Computer Security For Activists & Everyone (Oct 2018)
Kit O'Connell
 
Present Like a Pro | Leah Guren
Present Like a Pro | Leah GurenPresent Like a Pro | Leah Guren
Present Like a Pro | Leah Guren
LavaConConference
 
Unmasking miscreants
Unmasking miscreantsUnmasking miscreants
Unmasking miscreants
Brandon Levene
 
International Cooperative: APT Hunting
International Cooperative: APT HuntingInternational Cooperative: APT Hunting
International Cooperative: APT Hunting
Joshua Lawton, MBA
 
YOW2018 CTO Summit: Working at netflix
YOW2018 CTO Summit: Working at netflixYOW2018 CTO Summit: Working at netflix
YOW2018 CTO Summit: Working at netflix
Brendan Gregg
 
Securing your digital life - Jason Addie
Securing your digital life -  Jason AddieSecuring your digital life -  Jason Addie
Securing your digital life - Jason Addie
DataFest Tbilisi
 
Passive Intelligence Gathering and Analytics - It's All Just Metadata!
Passive Intelligence Gathering and Analytics - It's All Just Metadata!Passive Intelligence Gathering and Analytics - It's All Just Metadata!
Passive Intelligence Gathering and Analytics - It's All Just Metadata!
CTruncer
 
Understanding and implementing website security
Understanding and implementing website securityUnderstanding and implementing website security
Understanding and implementing website security
Drew Gorton
 
Seizing Electronic Evidence & Best Practices – Secret Service
Seizing Electronic Evidence & Best Practices – Secret ServiceSeizing Electronic Evidence & Best Practices – Secret Service
Seizing Electronic Evidence & Best Practices – Secret Service
Gol D Roger
 
State of the art in Natural Language Processing (March 2019)
State of the art in Natural Language Processing (March 2019)State of the art in Natural Language Processing (March 2019)
State of the art in Natural Language Processing (March 2019)
Liad Magen
 
Online Privacy & Computer Security Basics (September 2017)
Online Privacy & Computer Security Basics (September 2017)Online Privacy & Computer Security Basics (September 2017)
Online Privacy & Computer Security Basics (September 2017)
Kit O'Connell
 
A living hell - lessons learned in eight years of parsing real estate data
A living hell - lessons learned in eight years of parsing real estate data  A living hell - lessons learned in eight years of parsing real estate data
A living hell - lessons learned in eight years of parsing real estate data
lokku
 
Guerrilla UX: Practical and Affordable Research
Guerrilla UX: Practical and Affordable ResearchGuerrilla UX: Practical and Affordable Research
Guerrilla UX: Practical and Affordable Research
Brad Orego (he/they)
 
A DevOps Checklist for Startups
A DevOps Checklist for StartupsA DevOps Checklist for Startups
A DevOps Checklist for Startups
Rick Manelius
 
Байки із пожежного депо або як працює Big Data в Sigma Software, Денис Пишьєв,
Байки із пожежного депо або як працює Big Data в Sigma Software, Денис Пишьєв,Байки із пожежного депо або як працює Big Data в Sigma Software, Денис Пишьєв,
Байки із пожежного депо або як працює Big Data в Sigma Software, Денис Пишьєв,
Sigma Software
 
Hyper Island - 2012
Hyper Island - 2012Hyper Island - 2012
Hyper Island - 2012
Detectify
 
1.1. course introduction
1.1. course introduction1.1. course introduction
1.1. course introductionNicholas Wong
 
Internet Privacy
Internet PrivacyInternet Privacy
Internet Privacy
Girindro Pringgo Digdo
 
Privacy, security and ethics in data science
Privacy, security and ethics in data sciencePrivacy, security and ethics in data science
Privacy, security and ethics in data science
Nikolaos Vasiloglou
 
On hacking & security
On hacking & security On hacking & security
On hacking & security
Ange Albertini
 

Similar to Digital forensics and giving evidence by Jonathan Haddock (20)

Computer Security For Activists & Everyone (Oct 2018)
Computer Security For Activists & Everyone (Oct 2018)Computer Security For Activists & Everyone (Oct 2018)
Computer Security For Activists & Everyone (Oct 2018)
 
Present Like a Pro | Leah Guren
Present Like a Pro | Leah GurenPresent Like a Pro | Leah Guren
Present Like a Pro | Leah Guren
 
Unmasking miscreants
Unmasking miscreantsUnmasking miscreants
Unmasking miscreants
 
International Cooperative: APT Hunting
International Cooperative: APT HuntingInternational Cooperative: APT Hunting
International Cooperative: APT Hunting
 
YOW2018 CTO Summit: Working at netflix
YOW2018 CTO Summit: Working at netflixYOW2018 CTO Summit: Working at netflix
YOW2018 CTO Summit: Working at netflix
 
Securing your digital life - Jason Addie
Securing your digital life -  Jason AddieSecuring your digital life -  Jason Addie
Securing your digital life - Jason Addie
 
Passive Intelligence Gathering and Analytics - It's All Just Metadata!
Passive Intelligence Gathering and Analytics - It's All Just Metadata!Passive Intelligence Gathering and Analytics - It's All Just Metadata!
Passive Intelligence Gathering and Analytics - It's All Just Metadata!
 
Understanding and implementing website security
Understanding and implementing website securityUnderstanding and implementing website security
Understanding and implementing website security
 
Seizing Electronic Evidence & Best Practices – Secret Service
Seizing Electronic Evidence & Best Practices – Secret ServiceSeizing Electronic Evidence & Best Practices – Secret Service
Seizing Electronic Evidence & Best Practices – Secret Service
 
State of the art in Natural Language Processing (March 2019)
State of the art in Natural Language Processing (March 2019)State of the art in Natural Language Processing (March 2019)
State of the art in Natural Language Processing (March 2019)
 
Online Privacy & Computer Security Basics (September 2017)
Online Privacy & Computer Security Basics (September 2017)Online Privacy & Computer Security Basics (September 2017)
Online Privacy & Computer Security Basics (September 2017)
 
A living hell - lessons learned in eight years of parsing real estate data
A living hell - lessons learned in eight years of parsing real estate data  A living hell - lessons learned in eight years of parsing real estate data
A living hell - lessons learned in eight years of parsing real estate data
 
Guerrilla UX: Practical and Affordable Research
Guerrilla UX: Practical and Affordable ResearchGuerrilla UX: Practical and Affordable Research
Guerrilla UX: Practical and Affordable Research
 
A DevOps Checklist for Startups
A DevOps Checklist for StartupsA DevOps Checklist for Startups
A DevOps Checklist for Startups
 
Байки із пожежного депо або як працює Big Data в Sigma Software, Денис Пишьєв,
Байки із пожежного депо або як працює Big Data в Sigma Software, Денис Пишьєв,Байки із пожежного депо або як працює Big Data в Sigma Software, Денис Пишьєв,
Байки із пожежного депо або як працює Big Data в Sigma Software, Денис Пишьєв,
 
Hyper Island - 2012
Hyper Island - 2012Hyper Island - 2012
Hyper Island - 2012
 
1.1. course introduction
1.1. course introduction1.1. course introduction
1.1. course introduction
 
Internet Privacy
Internet PrivacyInternet Privacy
Internet Privacy
 
Privacy, security and ethics in data science
Privacy, security and ethics in data sciencePrivacy, security and ethics in data science
Privacy, security and ethics in data science
 
On hacking & security
On hacking & security On hacking & security
On hacking & security
 

More from Alex Cachia

No Onions, No Tiers - An Introduction to Vertical Slice Architecture by Bill ...
No Onions, No Tiers - An Introduction to Vertical Slice Architecture by Bill ...No Onions, No Tiers - An Introduction to Vertical Slice Architecture by Bill ...
No Onions, No Tiers - An Introduction to Vertical Slice Architecture by Bill ...
Alex Cachia
 
Supporting IT by David Meares
Supporting IT by David MearesSupporting IT by David Meares
Supporting IT by David Meares
Alex Cachia
 
OWASP Top 10 2021 - let's take a closer look by Glenn Wilson
OWASP Top 10 2021 - let's take a closer look by Glenn WilsonOWASP Top 10 2021 - let's take a closer look by Glenn Wilson
OWASP Top 10 2021 - let's take a closer look by Glenn Wilson
Alex Cachia
 
If you think open source is not for you, think again by Jane Chakravorty
If you think open source is not for you, think again by Jane ChakravortyIf you think open source is not for you, think again by Jane Chakravorty
If you think open source is not for you, think again by Jane Chakravorty
Alex Cachia
 
Chaos Engineering – why we should all practice breaking things on purpose by ...
Chaos Engineering – why we should all practice breaking things on purpose by ...Chaos Engineering – why we should all practice breaking things on purpose by ...
Chaos Engineering – why we should all practice breaking things on purpose by ...
Alex Cachia
 
A brief overview of the history and practice of user experience by Ian Westbrook
A brief overview of the history and practice of user experience by Ian WestbrookA brief overview of the history and practice of user experience by Ian Westbrook
A brief overview of the history and practice of user experience by Ian Westbrook
Alex Cachia
 
Return the carriage, feed the line by Aaron Taylor
Return the carriage, feed the line by Aaron TaylorReturn the carriage, feed the line by Aaron Taylor
Return the carriage, feed the line by Aaron Taylor
Alex Cachia
 
Treating your career path and training like leveling up in games by Raymond C...
Treating your career path and training like leveling up in games by Raymond C...Treating your career path and training like leveling up in games by Raymond C...
Treating your career path and training like leveling up in games by Raymond C...
Alex Cachia
 
Software Security by Glenn Wilson
Software Security by Glenn WilsonSoftware Security by Glenn Wilson
Software Security by Glenn Wilson
Alex Cachia
 
Data Preparation and the Importance of How Machines Learn by Rebecca Vickery
Data Preparation and the Importance of How Machines Learn by Rebecca VickeryData Preparation and the Importance of How Machines Learn by Rebecca Vickery
Data Preparation and the Importance of How Machines Learn by Rebecca Vickery
Alex Cachia
 
Why Rust? by Edd Barrett (codeHarbour December 2019)
Why Rust? by Edd Barrett (codeHarbour December 2019)Why Rust? by Edd Barrett (codeHarbour December 2019)
Why Rust? by Edd Barrett (codeHarbour December 2019)
Alex Cachia
 
Issue with tracking? Fail that build! by Steve Coppin-Smith (codeHarbour Nove...
Issue with tracking? Fail that build! by Steve Coppin-Smith (codeHarbour Nove...Issue with tracking? Fail that build! by Steve Coppin-Smith (codeHarbour Nove...
Issue with tracking? Fail that build! by Steve Coppin-Smith (codeHarbour Nove...
Alex Cachia
 
Hack your voicemail with Javascript by Chris Willmott (codeHarbour October 2019)
Hack your voicemail with Javascript by Chris Willmott (codeHarbour October 2019)Hack your voicemail with Javascript by Chris Willmott (codeHarbour October 2019)
Hack your voicemail with Javascript by Chris Willmott (codeHarbour October 2019)
Alex Cachia
 
Developing for Africa by Jonathan Haddock (codeHarbour October 2019)
Developing for Africa by Jonathan Haddock (codeHarbour October 2019)Developing for Africa by Jonathan Haddock (codeHarbour October 2019)
Developing for Africa by Jonathan Haddock (codeHarbour October 2019)
Alex Cachia
 
Revving up with Reinforcement Learning by Ricardo Sueiras
Revving up with Reinforcement Learning by Ricardo SueirasRevving up with Reinforcement Learning by Ricardo Sueiras
Revving up with Reinforcement Learning by Ricardo Sueiras
Alex Cachia
 
Blockchain For Your Business by Kenneth Cox (codeHarbour July 2019)
Blockchain For Your Business by Kenneth Cox (codeHarbour July 2019)Blockchain For Your Business by Kenneth Cox (codeHarbour July 2019)
Blockchain For Your Business by Kenneth Cox (codeHarbour July 2019)
Alex Cachia
 
Seeking Simplicity by Phil Nash (codeHarbour June 2019)
Seeking Simplicity by Phil Nash (codeHarbour June 2019)Seeking Simplicity by Phil Nash (codeHarbour June 2019)
Seeking Simplicity by Phil Nash (codeHarbour June 2019)
Alex Cachia
 
Sharing Data is Caring Data by Mark Terry (codeHarbour June 2019)
Sharing Data is Caring Data by Mark Terry (codeHarbour June 2019)Sharing Data is Caring Data by Mark Terry (codeHarbour June 2019)
Sharing Data is Caring Data by Mark Terry (codeHarbour June 2019)
Alex Cachia
 
Managing technical debt by Chris Willmott (codeHarbour April 2019)
Managing technical debt by Chris Willmott (codeHarbour April 2019)Managing technical debt by Chris Willmott (codeHarbour April 2019)
Managing technical debt by Chris Willmott (codeHarbour April 2019)
Alex Cachia
 
Telephone Systems and Voice over IP by Bob Eager (codeHarbour April 2019)
Telephone Systems and Voice over IP by Bob Eager (codeHarbour April 2019)Telephone Systems and Voice over IP by Bob Eager (codeHarbour April 2019)
Telephone Systems and Voice over IP by Bob Eager (codeHarbour April 2019)
Alex Cachia
 

More from Alex Cachia (20)

No Onions, No Tiers - An Introduction to Vertical Slice Architecture by Bill ...
No Onions, No Tiers - An Introduction to Vertical Slice Architecture by Bill ...No Onions, No Tiers - An Introduction to Vertical Slice Architecture by Bill ...
No Onions, No Tiers - An Introduction to Vertical Slice Architecture by Bill ...
 
Supporting IT by David Meares
Supporting IT by David MearesSupporting IT by David Meares
Supporting IT by David Meares
 
OWASP Top 10 2021 - let's take a closer look by Glenn Wilson
OWASP Top 10 2021 - let's take a closer look by Glenn WilsonOWASP Top 10 2021 - let's take a closer look by Glenn Wilson
OWASP Top 10 2021 - let's take a closer look by Glenn Wilson
 
If you think open source is not for you, think again by Jane Chakravorty
If you think open source is not for you, think again by Jane ChakravortyIf you think open source is not for you, think again by Jane Chakravorty
If you think open source is not for you, think again by Jane Chakravorty
 
Chaos Engineering – why we should all practice breaking things on purpose by ...
Chaos Engineering – why we should all practice breaking things on purpose by ...Chaos Engineering – why we should all practice breaking things on purpose by ...
Chaos Engineering – why we should all practice breaking things on purpose by ...
 
A brief overview of the history and practice of user experience by Ian Westbrook
A brief overview of the history and practice of user experience by Ian WestbrookA brief overview of the history and practice of user experience by Ian Westbrook
A brief overview of the history and practice of user experience by Ian Westbrook
 
Return the carriage, feed the line by Aaron Taylor
Return the carriage, feed the line by Aaron TaylorReturn the carriage, feed the line by Aaron Taylor
Return the carriage, feed the line by Aaron Taylor
 
Treating your career path and training like leveling up in games by Raymond C...
Treating your career path and training like leveling up in games by Raymond C...Treating your career path and training like leveling up in games by Raymond C...
Treating your career path and training like leveling up in games by Raymond C...
 
Software Security by Glenn Wilson
Software Security by Glenn WilsonSoftware Security by Glenn Wilson
Software Security by Glenn Wilson
 
Data Preparation and the Importance of How Machines Learn by Rebecca Vickery
Data Preparation and the Importance of How Machines Learn by Rebecca VickeryData Preparation and the Importance of How Machines Learn by Rebecca Vickery
Data Preparation and the Importance of How Machines Learn by Rebecca Vickery
 
Why Rust? by Edd Barrett (codeHarbour December 2019)
Why Rust? by Edd Barrett (codeHarbour December 2019)Why Rust? by Edd Barrett (codeHarbour December 2019)
Why Rust? by Edd Barrett (codeHarbour December 2019)
 
Issue with tracking? Fail that build! by Steve Coppin-Smith (codeHarbour Nove...
Issue with tracking? Fail that build! by Steve Coppin-Smith (codeHarbour Nove...Issue with tracking? Fail that build! by Steve Coppin-Smith (codeHarbour Nove...
Issue with tracking? Fail that build! by Steve Coppin-Smith (codeHarbour Nove...
 
Hack your voicemail with Javascript by Chris Willmott (codeHarbour October 2019)
Hack your voicemail with Javascript by Chris Willmott (codeHarbour October 2019)Hack your voicemail with Javascript by Chris Willmott (codeHarbour October 2019)
Hack your voicemail with Javascript by Chris Willmott (codeHarbour October 2019)
 
Developing for Africa by Jonathan Haddock (codeHarbour October 2019)
Developing for Africa by Jonathan Haddock (codeHarbour October 2019)Developing for Africa by Jonathan Haddock (codeHarbour October 2019)
Developing for Africa by Jonathan Haddock (codeHarbour October 2019)
 
Revving up with Reinforcement Learning by Ricardo Sueiras
Revving up with Reinforcement Learning by Ricardo SueirasRevving up with Reinforcement Learning by Ricardo Sueiras
Revving up with Reinforcement Learning by Ricardo Sueiras
 
Blockchain For Your Business by Kenneth Cox (codeHarbour July 2019)
Blockchain For Your Business by Kenneth Cox (codeHarbour July 2019)Blockchain For Your Business by Kenneth Cox (codeHarbour July 2019)
Blockchain For Your Business by Kenneth Cox (codeHarbour July 2019)
 
Seeking Simplicity by Phil Nash (codeHarbour June 2019)
Seeking Simplicity by Phil Nash (codeHarbour June 2019)Seeking Simplicity by Phil Nash (codeHarbour June 2019)
Seeking Simplicity by Phil Nash (codeHarbour June 2019)
 
Sharing Data is Caring Data by Mark Terry (codeHarbour June 2019)
Sharing Data is Caring Data by Mark Terry (codeHarbour June 2019)Sharing Data is Caring Data by Mark Terry (codeHarbour June 2019)
Sharing Data is Caring Data by Mark Terry (codeHarbour June 2019)
 
Managing technical debt by Chris Willmott (codeHarbour April 2019)
Managing technical debt by Chris Willmott (codeHarbour April 2019)Managing technical debt by Chris Willmott (codeHarbour April 2019)
Managing technical debt by Chris Willmott (codeHarbour April 2019)
 
Telephone Systems and Voice over IP by Bob Eager (codeHarbour April 2019)
Telephone Systems and Voice over IP by Bob Eager (codeHarbour April 2019)Telephone Systems and Voice over IP by Bob Eager (codeHarbour April 2019)
Telephone Systems and Voice over IP by Bob Eager (codeHarbour April 2019)
 

Recently uploaded

Accelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish CachingAccelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish Caching
Thijs Feryn
 
When stars align: studies in data quality, knowledge graphs, and machine lear...
When stars align: studies in data quality, knowledge graphs, and machine lear...When stars align: studies in data quality, knowledge graphs, and machine lear...
When stars align: studies in data quality, knowledge graphs, and machine lear...
Elena Simperl
 
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
UiPathCommunity
 
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Product School
 
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Jeffrey Haguewood
 
Securing your Kubernetes cluster_ a step-by-step guide to success !
Securing your Kubernetes cluster_ a step-by-step guide to success !Securing your Kubernetes cluster_ a step-by-step guide to success !
Securing your Kubernetes cluster_ a step-by-step guide to success !
KatiaHIMEUR1
 
Bits & Pixels using AI for Good.........
Bits & Pixels using AI for Good.........Bits & Pixels using AI for Good.........
Bits & Pixels using AI for Good.........
Alison B. Lowndes
 
Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*
Frank van Harmelen
 
The Future of Platform Engineering
The Future of Platform EngineeringThe Future of Platform Engineering
The Future of Platform Engineering
Jemma Hussein Allen
 
Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...
Product School
 
Transcript: Selling digital books in 2024: Insights from industry leaders - T...
Transcript: Selling digital books in 2024: Insights from industry leaders - T...Transcript: Selling digital books in 2024: Insights from industry leaders - T...
Transcript: Selling digital books in 2024: Insights from industry leaders - T...
BookNet Canada
 
FIDO Alliance Osaka Seminar: Overview.pdf
FIDO Alliance Osaka Seminar: Overview.pdfFIDO Alliance Osaka Seminar: Overview.pdf
FIDO Alliance Osaka Seminar: Overview.pdf
FIDO Alliance
 
Elevating Tactical DDD Patterns Through Object Calisthenics
Elevating Tactical DDD Patterns Through Object CalisthenicsElevating Tactical DDD Patterns Through Object Calisthenics
Elevating Tactical DDD Patterns Through Object Calisthenics
Dorra BARTAGUIZ
 
Connector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a buttonConnector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a button
DianaGray10
 
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdfFIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
FIDO Alliance
 
PCI PIN Basics Webinar from the Controlcase Team
PCI PIN Basics Webinar from the Controlcase TeamPCI PIN Basics Webinar from the Controlcase Team
PCI PIN Basics Webinar from the Controlcase Team
ControlCase
 
Monitoring Java Application Security with JDK Tools and JFR Events
Monitoring Java Application Security with JDK Tools and JFR EventsMonitoring Java Application Security with JDK Tools and JFR Events
Monitoring Java Application Security with JDK Tools and JFR Events
Ana-Maria Mihalceanu
 
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMsTo Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
Paul Groth
 
How world-class product teams are winning in the AI era by CEO and Founder, P...
How world-class product teams are winning in the AI era by CEO and Founder, P...How world-class product teams are winning in the AI era by CEO and Founder, P...
How world-class product teams are winning in the AI era by CEO and Founder, P...
Product School
 
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
Product School
 

Recently uploaded (20)

Accelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish CachingAccelerate your Kubernetes clusters with Varnish Caching
Accelerate your Kubernetes clusters with Varnish Caching
 
When stars align: studies in data quality, knowledge graphs, and machine lear...
When stars align: studies in data quality, knowledge graphs, and machine lear...When stars align: studies in data quality, knowledge graphs, and machine lear...
When stars align: studies in data quality, knowledge graphs, and machine lear...
 
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
 
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...
 
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
 
Securing your Kubernetes cluster_ a step-by-step guide to success !
Securing your Kubernetes cluster_ a step-by-step guide to success !Securing your Kubernetes cluster_ a step-by-step guide to success !
Securing your Kubernetes cluster_ a step-by-step guide to success !
 
Bits & Pixels using AI for Good.........
Bits & Pixels using AI for Good.........Bits & Pixels using AI for Good.........
Bits & Pixels using AI for Good.........
 
Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*Neuro-symbolic is not enough, we need neuro-*semantic*
Neuro-symbolic is not enough, we need neuro-*semantic*
 
The Future of Platform Engineering
The Future of Platform EngineeringThe Future of Platform Engineering
The Future of Platform Engineering
 
Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...
 
Transcript: Selling digital books in 2024: Insights from industry leaders - T...
Transcript: Selling digital books in 2024: Insights from industry leaders - T...Transcript: Selling digital books in 2024: Insights from industry leaders - T...
Transcript: Selling digital books in 2024: Insights from industry leaders - T...
 
FIDO Alliance Osaka Seminar: Overview.pdf
FIDO Alliance Osaka Seminar: Overview.pdfFIDO Alliance Osaka Seminar: Overview.pdf
FIDO Alliance Osaka Seminar: Overview.pdf
 
Elevating Tactical DDD Patterns Through Object Calisthenics
Elevating Tactical DDD Patterns Through Object CalisthenicsElevating Tactical DDD Patterns Through Object Calisthenics
Elevating Tactical DDD Patterns Through Object Calisthenics
 
Connector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a buttonConnector Corner: Automate dynamic content and events by pushing a button
Connector Corner: Automate dynamic content and events by pushing a button
 
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdfFIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
FIDO Alliance Osaka Seminar: Passkeys and the Road Ahead.pdf
 
PCI PIN Basics Webinar from the Controlcase Team
PCI PIN Basics Webinar from the Controlcase TeamPCI PIN Basics Webinar from the Controlcase Team
PCI PIN Basics Webinar from the Controlcase Team
 
Monitoring Java Application Security with JDK Tools and JFR Events
Monitoring Java Application Security with JDK Tools and JFR EventsMonitoring Java Application Security with JDK Tools and JFR Events
Monitoring Java Application Security with JDK Tools and JFR Events
 
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMsTo Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
 
How world-class product teams are winning in the AI era by CEO and Founder, P...
How world-class product teams are winning in the AI era by CEO and Founder, P...How world-class product teams are winning in the AI era by CEO and Founder, P...
How world-class product teams are winning in the AI era by CEO and Founder, P...
 
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
 

Digital forensics and giving evidence by Jonathan Haddock

  • 1. Digital forensics and giving evidence With Jonathan Haddock
  • 2. Show and tell - digital forensics and giving evidence Jonathan Haddock March 2020
  • 3. About me ● I’ve worked in IT for over 15 years ● Now the Deputy Senior Information Security Officer for three local authorities ● Have run penetration testing courses with the YPISG ● Trained in forensics, penetration testing and incident response https://keybase.io/joncojonathan @joncojonathan https://uk.linkedin.com/in/jonathanhaddock
  • 4. What is digital forensics? ● Process of looking at computer systems and files to determine what activity took place ○ Can also be about recovery ● Increasingly important for incident response ● Not like it's shown in the movies! ○ You can't decrypt part of a drive like that ● Isn't “wet” forensics - we're in the digital realm ● Needs translation and explanation
  • 5. Why translation? ● Logs can be really hard to read (or just plain horrible!) This just means someone accessed a file on a site, and got redirected (resource permanently moved)... 188.65.57.4 - - [09/Oct/2018:20:32:40 +0000] "GET /RSS/blogfeed.xml?blogID=1 HTTP/1.1" 301 185 "http://example.org/RSS/blogfeed.xml?blogID=1" "SimplePie/1.2 (Feed Parser; http://simplepie.org; Allow like Gecko) Build/20090627192103"
  • 6. What digital forensics isn't ● Not an instant indicator of blame ○ Also not your fault ● (Often) Unable to conclusively tie actions to a human being ○ Unless there's a camera pointing at the user and the screen, you can't know for certain who did what ● Not as easy as it used to be thanks to encryption
  • 7. Forensic process ● Obtain the assets ○ Files, disk images or physical media, phones ● Calculate the asset’s hash ● Take an image of the asset ● Calculate hashes ● Make a copy ● Check its hash ● Perform analysis
  • 8. Chain of custody ● First entry: Assets received by you from provider ● Shows who had access, so possibility of tampering reduced ● If the asset moves, is copied or shown to someone else, note in the log ● Protects you
  • 9. Taking hashes ● A hash is a mathematically calculated signature for a file ○ da39a3ee5e6b4b0d3255bfef95601890afd80709 ● md5, sha1, sha2 ● Note tool version in your report ○ sha1sum (GNU coreutils) 8.29 ● The sum will change as soon as the file contents is edited
  • 10. Taking an image ● Not just a photograph! ● A block by block copy of a disk ○ Tools like dd or ftkimager ○ Not a copy of the files on the disk ● Means we can leave the original untouched ● Beware of Windows - it’s not forensically safe ○ Use a write blocker
  • 12. Make a copy of the image ● Immediately after the original is received or the disk image created ○ Making disk images takes time, you don’t want to do it again ● Check the hash ● Move the original to protected storage ● We’ll work on a copy ● You may have to provide a copy
  • 13. Forensic tools ● Tools like Encase or Forensic Tool Kit (FTK) are expensive - around $3,000 each ● The Sleuth Kit is open source ○ Has a graphical front end - Autopsy ● PhotoRec (free) can recover files ● But sometimes you just have to use your everyday applications
  • 14.
  • 15. Analysing the evidence ● Pick the relevant tool ○ Cost ○ Task ● Stick to the specification ○ If looking for files or events in one date range, don’t get distracted by others ● Try to avoid changing the copy - avoids confusion ● Take notes / keep a journal
  • 16. Writing reports and statements ● Take information from your journal / notes ● Write for the audience - probably not too technical ○ Internal or external? ● Appendix: ○ List of hashes ○ Maybe chain of custody ● Also need-to-know
  • 17. Why “need to know”? ● We have a duty of care to everyone involved in the case ○ Prosecution ○ The accused ● Leaked details can still ruin someone’s life ○ Newspapers often publish details before court judgments ● UK law: innocent until proven guilty
  • 18. Court
  • 19. Attending court ● You’re not allowed in before giving your evidence ● Expect to sit around a lot ○ Courtroom might be triple booked ● Look at your questioner ● Answer to the Magistrates ○ Watch their pens - don’t talk too fast ● Make sure your phone is off!
  • 20. Be clear, explain ● Magistrates are lay people ● Your job is to explain what the evidence means ● You might be allowed a copy of your statement ○ Can’t be annotated ● If you don’t understand the question, get it repeated or rephrased
  • 21. Don’t make stuff up! ● You either found it in the evidence, or you didn’t ● If you don’t know, say you don’t know ○ If you can’t remember, say that too ● Answers should clarify or explain your statement ○ You’re not creating new information, just explaining what’s already presented
  • 22. Cross examination ● The opposition barrister may want to ask questions ● Their aim is to help their client, don’t feel attacked ○ Consider their questions as just other questions ● You may be asked for clarification to your earlier answers ● Alternatively, there may be entirely new questions ● The first barrister may ask additional questions
  • 23. Assumption trap ● “Is it fair to assume that…” ● Beware - is it fair to assume anything in criminal proceedings? ● Ultimately we’re not there to speculate ● Can be a way to use your expertise to cause you to lend support to a tenuous idea
  • 24. Reporters ● There’s likely to be reporters somewhere ○ Courtroom (public gallery) ○ Waiting area (public space) ○ Just outside ○ May be in disguise ● Don’t be afraid to say “no comment” ● They can be pushy
  • 27. Interested in security? Try a podcast Smashing Security Humorous run down of security news for the week. With Graham Cluley (@gcluley) and Carole Theriault (@caroletheriault). www.smashingsecurity.com Security Now A semi-technical dive into recent security news and vulnerabilities. Hosted by Leo Laporte and Steve Gibson (@sggrc). www.twit.tv/sn
  • 28. The slides that follow were not presented at codeHarbour due to time, but give you the rest of the story. I also blogged about this at https://blog.jonsdocs.org.uk/2018/11/12/show-and-tell-digital- forensics-and-giving-evidence/
  • 29. Hash example ● Simple text file, says “hello” f572d396fae9206628714fb2ce00f72e94f2258f ● Text file edited to say “hell0” 40dfb682064a88af16ddb3faa2d3b5f0e6a82ee3 ● A minor change has completely changed the hash
  • 30. Validate your tools ● As tools receive updates their output can change ● Have a known sample ● After an upgrade, confirm the tool still gives known results ○ If file spoons.jpg was found before, it still should be, and be identical ● New versions may add new features, more results
  • 31. Keeping track ● I always keep a journal ○ Details of assignment ○ Findings (each artefact has a hash) ○ Screenshots ○ Thoughts ○ Chain of custody ● Journal should be shared on a need-to-know basis
  • 32. References and acknowledgements ● Screenshots of the San Bernardino case reporting: ○ https://www.theguardian.com/technology/2016/feb/22/tim-cook-apple- refusal-unlock-iphone-fbi-civil-liberties ○ https://www.bbc.co.uk/news/technology-35593048 ● USB write blocker image: https://www.cru- inc.com/products/wiebetech/usb_writeblocker/ ● Autopsy screenshot from https://www.sleuthkit.org/autopsy/images/v3/overview.png ● Podcast logos © their respective owners, from their respective sites

Editor's Notes

  1. This presentation was originally written after some forensics work I did for the Information Commissioner’s Office (ICO). I started that work in 2017 and ended up in court giving evidence in 2018. Please note that while there are a number of products linked to in this presentation, I’m not on commission for them nor do I have any investment in them. It is up to the reader to determine if a product meets their needs.
  2. The YPISG is a BCS specialist group - Young Professionals’ Information Security Group. Every role helps with security, whether that‘s on the service desk or as a developer. I‘ve worked across a number of roles, starting in service desk / desktop support, and each has benefited from experience gained in the others. If you want to get into IT security, I recommend you specialise in an area you find interesting but make sure you keep your hand in other areas too so you have a broad skill set and knowledge. This will allow you to have informed discussions with different teams, in a way they (and you) can understand. I also blog about cyber security, development, IT, and occasionally local history at https://blog.jonsdocs.org.uk.
  3. Digital forensics ranges from the simple recovery of photos from a wiped SD card to a full investigation into the actions undertaken on a computer system. There’s an increased need for forensics awareness for incident response: determining how / where a cyber security incident (e.g. a malware infection or breach) occurred will allow you or others to contain the issue and prevent it happening again. In the movies there’s often a race against time to crack into the enemy’s encrypted hard disk - note you can’t usually break part of the key. Similarly, we can’t enhance photos from poor pixelated imagery to HD level quality! As this isn’t “wet forensics” there’s no DNA here, but we do have a number of dvantages as I’ll cover later.
  4. Part of the forensic analyst’s job is to translate what they’ve found into something that can be understood. Logs in particular can be horrible to review and a judge / magistrate / HR manager is unlikely to understand the raw evidence you find. This is an excerpt from a Nginx web server log. It’s fairly unintelligible unless you know what you’re looking at (Source IP, Date / time, HTTP method, Path, HTTP status code, URL, User agent, to name a few of the fields). Clearly we can’t just present the logs to someone, we need to explain them.
  5. Because this isn’t “wet forensics” it’s virtually impossible to concretely say “this particular human being performed this particular action” - you’d need a camera looking at the keyboard, mouse, monitor and user in order to do that. You can, and I’d suggest should, state in your report that your evidence identifies a user account or device, not a human being. Digital forensic evidence can act as part of the overall evidence bundle, but shouldn’t be used on its own to convict someone. When performing an investigation you might find yourself thinking “I’m going to cost this person their job, my report is quite damning”. This is a dangerous mindset that’s easy to fall in to. What you need to remember is that you aren’t going to cause someone to lose their job / go to jail, their actions have lead to this point (if they’re guilty). It’s important to be objective at all times and record your findings. Explain them but don’t attach your own interpretation to them - that isn’t your job as the analyst. Don’t forget we’re not creating evidence, we’re only reporting what’s there. Encryption has made investigation significantly harder, but it is good for privacy - a moral dilemma.
  6. If presented with files or pre-made images, it’s wise to enquire about their provenance. How were they made, by whom, when? Is there a reason for not being provided the physical media / device? We calculate a hash of the asset as this will describe the absolute original (this was originally omitted from the presentation, added during question time, and has been added here for completeness). Hashing is covered later in this presentation. By taking an image of the asset we have a file we can perform our analysis on, protecting the original from changes. This is a great advantage over “wet forensics”. Consider the example of this room - with all of us having come to the event, any “wet forensic” evidence that was already in here has been damaged or destroyed by our presence. In digital forensics we can always refer to the original image, so our environment, so to speak, can always be pristine. Every time the file moves we calculate the hash again, just to make sure the copy is valid.
  7. This document underpins all our work, and helps protect us / the case from allegations of tampering. As the log shows who could access the evidence it’s possible to determine everyone that could have made changes or provided an interpretation.
  8. By taking or calculating a hash you determine a fingerprint for a file or device. As shown on the next slide, any change will produce a change to the hash. After each file copy we calculate the hash of the copy to prove the files are the same. Collisions are unlikely but can happen for MD5 and SHA1. A collision is when two different inputs provide the same hash. I used MD5 and SHA1 during my investigation as finding two different inputs that provide the same hash for both algorithms seems infeasible, especially when you consider the input would still have to be relevant to the case.
  9. The difference between taking a copy of the files on physical media and making a bit by bit copy is the former only shows the current state of affairs. When a file is deleted, the file system (NTFS, EXT4, FAT32 etc.) simply marks it in the table as no longer being present. Space is also returned to the operating system so a new file could be placed where the old one was - but really the file is still there. In a bit by bit copy of the media we have a copy of the whole file system, not just the present files, so we can recover deleted files through a process known as data carving. Microsoft Windows automatically mounts (makes available for use) disks and in doing so places a marker on the file system, thus changing the original media. Linux doesn’t (always) do this, but in either case can be instructed to mount media read only (mount -o ro). If doing forensics on Windows it’s important to use a write blocker (like the USB one, pictured) which intercepts and discards any instructions to write to the file system.
  10. As mentioned previously, this is about protecting the original. After we’ve taken our image of the original asset it should be locked away, and noted in the chain of custody as such, with all work performed on a copy of the image. Why on a copy? Images can take a long time to produce, and you don’t want to repeat that exercise (especially given a lot of forensics is performed in environments where cost is very much a factor, and time is money). You may be asked to provide a copy to expert witnesses or the opposition, so they can perform their own analysis.
  11. It’s unlikely you’ll have FTK (https://accessdata.com/products-services/forensic-toolkit-ftk) or Encase (https://www.guidancesoftware.com/) to hand, as they’re used by law enforcement or consultancies that undertake forensic work regularly. They are powerful tools, but for casual or irregular use perhaps a poor investment. The Sleuth Kit (http://www.sleuthkit.org/) and Autopsy (http://www.sleuthkit.org/autopsy/) are free and becoming increasingly powerful. Initially Autopsy was a web based front end but as you’ll see in the screenshot later on it’s changed to be an application now. Autopsy supports a number of features, including the ability to look at at browser histories. Photorec by CG Security (https://www.cgsecurity.org/wiki/PhotoRec) is another free tool that started out as a data carver to recover photos, but now supports recovery of multiple file formats. In my case I was dealing with PST file exports from a Microsoft Exchange server. Microsoft Outlook was my tool of choice as it’s designed to work with these files, although I still followed forensic practice.
  12. A screenshot of Autopsy, from its website. We can see it’s managed to pull information from the browser (cookies, history etc.) and Autopsy has categorised files for easier analysis.
  13. It’s very easy to get distracted by your findings. In my case, I was reviewing a mailbox that contained thousands of emails, the vast majority of which were of no relevance to the investigation I was undertaking. I had no need, or right, to review those emails. To avoid distraction, stick to the brief: if you’re looking for changes in a certain time frame, restrict yourself to it. You may find items relating to the same topic outside the time range, but they’re outside of scope. Avoid making changes wherever possible, even down to rotating pictures or renaming files in the image as changes can lead to confusion: did I do that, or was the file already like that? While we have the option of referring to a pristine copy, it’s not desirable to have to keep looking at fresh copies as this takes time. Each item we extract from the image (called an artefact), in my case emails, should have a hash calculated and this should be noted.
  14. Your report or statement has to be accurate, that’s the most important thing. Don’t make stuff up, do your best to avoid errors, and don’t be ambiguous. I read my statements through several times before I signed them (in my case, every single page needed my signature). It may be appropriate to include your chain of custody, dependant on your audience. The level of technical detail needs to be appropriate. Don’t go into extensive detail on why data carving works - the reader isn’t likely to need that information. Instead, adopt a style that allows you to remain brief and factual. You shouldn’t be making any assumptions in your report - the facts will speak for themselves. Refer to your journal for details to include in the report. For me, the intended audience was barristers and eventually magistrates in court.
  15. You’ll note in the press that quite often there are full page spreads and photographs devoted to “Joe Bloggs, 37, arrested for abuse of…” and similar. This is dangerous as it immediately biases the public and can actually destroy lives. I knew a teacher in my secondary school who was falsely accused (the accuser made up the allegation) - details were spread of the accusation and the teacher had to leave the area, eventually leaving teaching. Their life was destroyed due to a lack of “need to know”. UK law says an individual is innocent until proven guilty, but leaking details can mean the public have already made up their mind regardless of the trial outcome. Be fair to the accused, keep your findings to yourself.
  16. I was at a magistrates’ court, meaning I was presenting evidence to three people that were acting like judges. They don’t have a legal background, instead taking advice from a legal advisor. My biggest worry was that I’d be discredited by the defence, like you see in the movies. My second biggest worry was that I had no idea what to expect in terms of questions. The barrister on your side (prosecution in my case) isn’t allowed to tell you the questions in advance as that’s known as “coaching a witness”. Initially I wasn’t allowed my statement for reference either, although the magistrates approved for me to receive this once I’d started giving evidence. Strangely, I wasn’t requested to provide any ID on arrival at court. I suspect this is because on first entering the witness box you’re asked to swear an oath to tell the truth, and then state your full name. Lying under oath is committing perjury and has legal ramifications. The magistrates may be taking notes - if so, watch their pens. Don’t talk too fast as if you’re too quick they won’t be able to make notes (or will stop listening in order to do so). Incidentally, you address the group of magistrates as “your worships”.
  17. In my case my technical evidence was not disputed, my attendance at court was to explain what it meant in a manner non-technical folk could understand. I could refer to my statement but importantly this wasn’t annotated, so any dates I needed to confirm I had to find at the time. To be fair to the defendant, I made sure to explain my investigation couldn’t pinpoint a particular human being. In the UK we have the concept of being innocent until proven guilty and given my job as a witness is to assist justice, and present the truth, it was only fair to highlight that.
  18. Remember - you’re not on trial, but the court is interested to hear what you have to say. That said, if you make something up, or lie, you are committing an offence. As you’re discussing your statement you don’t necessarily need to worry about recalling details you’ve left out although you may be asked to explain part of the process. You either found something that answers the question, or you didn’t. You’re not in court to create new information.
  19. Cross examination worried me but wasn’t a big deal. The questions were, literally, just other questions. Some questions I was asked were clearly intentionally ambiguous, so I ensured my answers were clear and precise. After cross examination the magistrates asked me further questions. Treat them exactly the same: with respect.
  20. I’d been warned about assumption questions by my employer’s legal team - be careful as this type of question can cause you to back yourself into a corner by making you think on your feet. The important thing is that it’s assumption. I was asked “is it fair to assume that because X happened on this date, that it’s the reason for Y”. I started looking at the dates, which matched, before realising this was an assumption trap. I pointed out that it wasn’t fair to assume anything in legal proceedings, as I had to be fair to the defendant and thus had to comment on my evidence.
  21. I have no problem with the free press, so long as they’re being honourable and accurate. During my time at court there were some reporters in the waiting area (they sat behind me), in the public gallery and outside. The one outside was in disguise (holding an unlit cigarette, I assumed he was going to ask me for a light). Don’t be afraid to say “no comment”. As the magistrates pointed out to me before I left the courtroom, I wasn’t to talk about the case to anyone outside (hence I’ve not gone into details here). If representing your employer they may have their own policy on speaking to the media.
  22. Smashing Security is the less technical of the two, with episodes lasting around 45 minutes. Most episodes there’s also a guest. Security Now episodes tend to last around two hours.
  23. Example showing the SHA1 hash for a text file containing the text “hello”, and how it changes completely following the edit.
  24. When a tool receives an update it’s possible that it will behave differently. To determine this it’s important to run tests against a known sample. Getting more information is not a problem, as the product may have simply added new features. No longer finding information would be a problem, and should be addressed through the vendor’s support channels.
  25. Your journal is not what you present as your evidence, it’s simply your working document. I record thoughts and to-do lists in mine, along with keeping a log of my chain of custody. Some forensic software allows you to record your notes in it, providing you a journal facility. If there were multiple researchers then the chain of custody would need to be a separate document. How you structure your journal is up to you - whatever works best. When working on a forensic investigation my journal contains details of who authorised the investigation, a copy of the specification and many tables of data along with screenshots of assets. Conversely, when I’m performing a penetration test (ethical hacking) my journal looks different and includes spider diagrams. There are some common elements between the two, but I change based on my task. Do whatever works best for you (if that’s pen and paper that’s fine, and I do that too, but remember to keep the physical copies secured).