Here is a thought experiment for you. A laptop is suspected of being used to copy confidential files. It is switched on, on a desk, and you are standing in front of it. What is the very first thing you do? If your instinct says “switch it off, to stop the damage”, pause, because that single action destroys some of the best evidence on the machine. Forensics begins with understanding that evidence has a lifespan.

Evidence evaporates at different speeds

Some data lasts for nanoseconds, some for years. The guideline for handling this is called the order of volatility, and it is written down in RFC 3227, a document on collecting and archiving evidence. You collect the most fleeting items first. Roughly, from most to least volatile:

  1. CPU registers and cache. In practice you almost never capture these.
  2. Network state: routing and ARP tables, active connections, running processes.
  3. Main memory (RAM): running programs, decrypted data, encryption keys, network sessions, sometimes malware that never touches the disk.
  4. Temporary files and swap space.
  5. Disk contents.
  6. Remote logging and monitoring data relevant to the system.
  7. Physical configuration and network topology.
  8. Archival media such as backups and tapes.

Why does switching off hurt? Power loss clears RAM. If the suspect’s laptop had a disk-encryption key loaded, or malware living only in memory, or a list of live connections to an outside server, it is gone. On the other hand, leaving the machine on carries risk: it may keep overwriting data, or someone may remotely wipe it. So responders weigh the two, and the decision should be documented with a reason. For a seized laptop that is already off, you simply leave it off and image the disk.

Work on copies, and prove the copy is faithful

The core principle: never analyse the original. You create a forensic image, a bit-for-bit copy of the storage, using a hardware write blocker or equivalent software controls so that connecting the drive does not alter it. Then you compute a cryptographic hash of the original and of the image (SHA-256 is common; older cases often used MD5, which is fine for detecting accidental change but weak against deliberate collision). If the two hashes match, the image is a faithful copy. You analyse the copy, and hash it again later to show it never changed.

The hash is a very short fingerprint of a very large thing. Change a single bit anywhere in a 500 GB image and the hash changes completely. That is what makes it powerful as proof of integrity.

Chain of custody: the paper trail

Technical integrity is half the story. The other half is being able to say who held the evidence, when, and why, from the moment of collection to the day it is presented. That record is the chain of custody. A good one lists:

  • What the item is (make, model, serial number, description, photographs of its condition).
  • Who collected it, from where, and at what date and time (state the time zone).
  • Each transfer: from whom, to whom, when, for what purpose.
  • How it was stored: sealed bags, tamper-evident labels, a locked cabinet with restricted access.
  • Hash values of images and when they were verified.

If there is a gap, such as an unlabelled drive left on someone’s desk overnight, the opposing side can argue it was tampered with. The evidence may still exist, but its credibility is damaged.

A word on Indian courts

Electronic records are admissible in India subject to conditions. Under the older Indian Evidence Act this was section 65B; since the Bharatiya Sakshya Adhiniyam, 2023 came into force in July 2024, the corresponding provision is section 63, which also requires a certificate accompanying electronic evidence, and my understanding is that the prescribed certificate format asks for details such as hash values. Check the current form with your legal counsel; the practical lesson for technical staff is to record device details, collection method and hashes carefully, because you may be asked to sign such a certificate later.

What forensic analysis looks for

Once you hold a trustworthy image, the questions are about reconstruction. Which files were created, modified or accessed, and when (file system timestamps)? Which programs ran (execution artefacts such as prefetch records on Windows)? What was plugged in (USB history)? What did the user browse, search or download? What was deleted? Deleted does not mean erased: when a file is deleted, the file system typically marks the space reusable but leaves the data until something overwrites it, which is why carving files out of unallocated space often works. On modern solid-state drives and with encryption in place, recovery is much less reliable.

Timelines are the main product. Merge file system times, event logs, browser history and application logs into one ordered sequence and the story emerges. Be careful with clocks: a machine’s clock may be wrong or have been changed, and time zones are an eternal source of off-by-five-and-a-half-hours mistakes for Indian teams working with UTC logs.

Habits that separate forensics from tinkering

Write down what you do while you do it, not afterwards. Stay inside your authority: collecting data from an employee’s personal phone or private accounts raises legal and privacy problems, so involve legal and HR first. Keep the scope to what the investigation needs. And report findings as facts you can support, separated from your interpretation, so that a second examiner repeating your steps reaches the same result. Repeatability is the quiet test of good forensic work.

Leave a Reply

Trending