Black Friday Skill Fest
BUY 1 GET 1 FREE
Limited-Time Offer

CISSP Domain 2 Cheat Sheet: Asset Security Concepts and Exam Tips

Comprehensive Last-Minute Revision Notes 

Aligned to the current ISC² CISSP Common Body of Knowledge 

Exam weight: ~10% 

Edition: October 2026 

How to use this?

Domain 2 is one of the smaller domains by weight, but it’s where well-prepared candidates quietly bank marks and where half-preparation gets punished. It’s full of crisp, almost technical distinctions: which sanitisation method suits which media, the three states of data, anonymisation versus pseudonymisation. Know them cold and they’re gifts; half-know them and they’re traps. 

The spine of the domain is simple: you cannot protect what you haven’t identified and valued, and protection must follow the asset all the way to the moment you destroy it. Everything here flows along that line: inventory, classification, handling, lifecycle, and disposal. 

Keep the manager’s altitude from Domain 1, but expect sharper recall here. Read it through once, then drill the Exam Alert boxes, the confused-pairs table, and the sanitisation rules the night before; those are where the easy marks live. 

MANAGER’S LENS: A theme to carry throughout: hoarding data and keeping dead assets ‘just in case’ is not caution; it’s accumulated liability. The least risky data is the data you never collected or have already securely destroyed.

1. The asset-security mindset

Asset security is the discipline of knowing what you own, understanding what it’s worth, and protecting it in proportion across its entire life; from the day you acquire it to the day you destroy it. Two ideas sit underneath every topic in this domain: 

  • Protection proportional to value: Don’t gold-plate public brochures; don’t under-protect the crown jewels. The level of control should track the sensitivity and criticality of the asset, judged by the impact if it’s disclosed, altered, or lost. 
  • Security spans the whole lifecycle; including disposal: A startling share of breaches come from the last step: a decommissioned drive sold on, a backup tape binned, a cloud bucket left behind. If your controls stop before destruction, they don’t really cover the asset. 

EXAM ALERT: The first step in securing assets, and the first step in classifying them is identifying and inventorying what you have. You cannot protect what you cannot see. 

2. Asset inventory and management

Everything starts with an accurate inventory. Assets come in two flavours: 

  • Tangible: Hardware, facilities, storage media, mobile devices. 
  • Intangible: Data, software and licences, intellectual property, brand and reputation, and the knowledge in people’s heads. 

For IT estates, a configuration management database (CMDB) records assets and the relationships between them, so you know what depends on what. Each asset moves through a lifecycle:

Acquisition → deployment and provisioning → operation and maintenance → decommissioning → disposal

Security requirements attach at every stage. 

Crucially, every asset needs an owner. Unowned assets don’t get classified, patched, monitored, or retired; they drift into shadow IT and become the weak point no one is watching. 

MANAGER’S LENS: An inventory isn’t a compliance checkbox; it’s the map your entire programme is drawn on. Unknown assets can’t be risk-assessed, so an incomplete inventory quietly caps how good your security can ever be. 

3. Information and data classification

Classification exists to match the level of protection to the value of the information efficiently. Over classifying wastes money and slows the business; under-classifying leaves sensitive data exposed. Both are failures. 

Data is classified against criteria such as its value, sensitivity, criticality, legal or regulatory obligations, and useful life (information often loses sensitivity over time, which is what declassification addresses). The data owner; a business role, not IT, decides the classification. 

Two schemes you should recognise instantly: 

Sector  Levels (low → high)
Commercial / private  Public → Sensitive → Private → Confidential (sometimes Proprietary)
Government / military  Unclassified → Sensitive But Unclassified (SBU) / Controlled Unclassified (CUI) → Confidential → Secret → Top Secret

The higher the classification, the stricter the handling, the fewer the people with access, and the stronger the controls. Classification should reflect the most sensitive element in a dataset and the impact of its compromise; not an average, and not a gut feeling. 

EXAM ALERT: The data owner assigns the classification; the data custodian applies the matching controls. Classify to the highest-sensitive data present, and tie the decision to business impact (CIA), not to how ‘secret’ something feels. 

CISSP

TRAINING CALENDAR of Upcoming Batches For CISSP Training

Start Date End Date Start - End Time Batch Type Training Mode Batch Status
07-Dec-2026 29-Dec-2026 07:00 - 12:00 IST Weekday Online [ Close ]
12-Dec-2026 24-Jan-2027 10:00 - 14:00 IST Weekend Online [ Open ]
09-Jan-2027 14-Feb-2027 19:00 - 23:00 IST Weekend Online [ Open ]
13-Feb-2027 21-Mar-2027 09:00 - 13:00 IST Weekend Online [ Open ]
13-Mar-2027 18-Apr-2027 19:00 - 23:00 IST Weekend Online [ Open ]

4. Asset and data roles

The same cast from Domain 1 reappears here, now viewed through handling and disposal. Keep the accountability-versus-implementation split clear. 

Role  Responsibility
Data / information owner  Classifies the data; accountable for its protection and acceptable use
System owner  Accountable for a specific system and its security posture
Business / mission owner  Owns the business process the data serves
Data controller  Determines the purpose and means of processing personal data
Data processor  Processes personal data on the controller’s instructions
Data custodian  Implements and maintains controls: storage, backups, retention, protection
Data steward  Responsible for data quality, context, and business meaning
Data subject  The individual the personal data is about (a person, not a job role)
Data user  Accesses and uses data in line with policy

EXAM ALERT: Owner sets, custodian maintains. The exam returns to this constantly. And the data subject is the person the data describes, not someone who handles it. 

5. Handling requirements: marking, storing, transmitting

Once data is classified, it has to be handled consistently at its level for the whole time you hold it. The handling requirements fall into a few buckets: 

  • Marking and labelling: Physical labels on media and digital tags or metadata on files announce the classification so people and systems know how to treat the asset. Mark both the physical and the electronic. 
  • Handling: Who may access it, how it’s transported, and who signs for custody. 
  • Storage: At the right protection level: encrypted at rest, locked media, access-controlled repositories. 
  • Transmission: Protected in transit with encryption whenever it crosses a network. 
  • Declassification: Formally lowering the classification (and re-marking) once the data no longer warrants the higher level. 

EXAM ALERT: Marking is what makes correct handling possible. Unlabelled media gets treated as low sensitivity by default and leaks; so ‘unlabelled’ must never be assumed to mean ‘unclassified’. When in doubt, handle at the higher level until it’s marked. 

6. The data lifecycle

Data has a life of its own: create or collect → store → use → share → archive and retain → destroy. Protection applies at every stage, and two stages catch people out. 

Data location, residency, and sovereignty

Where data physically lives determines which laws apply to it; the transborder-flow point from Domain 1, seen from the asset side. Cloud and SaaS make location slippery, so you have to know (and often contractually pin down) where your data and its backups actually sit. 

Data retention

Keep data only as long as there’s a genuine business or legal reason, then dispose of it. Retention schedules balance regulatory minimums (keep records for X years) against privacy and minimisation (don’t keep personal data forever). A legal hold overrides routine deletion when litigation is anticipated or underway. 

Get insights from the blog: What is a Data Retention policy?

MANAGER’S LENS: ‘Keep everything forever’ feels safe and is the opposite. Every extra record is more to breach, more to produce in discovery, and more regulatory exposure. Data minimisation; collecting less and deleting on schedule; is itself a security control. 

Preparing for CISSP?

Cover all eight domains with expert-led training. Explore InfosecTrain’s CISSP Certification Training.

7. Data states and how to protect each

At any moment, a piece of data is in one of three states, and each calls for a different kind of protection. Matching the control to the state is a classic exam move. 

State  What it means  How you protect it
Data at rest  Stored on disk, media, or in a database Encryption (AES, full-disk / file / database), access controls, physical security
Data in transit  Moving across a network  TLS, VPN / IPsec, SFTP, encrypted email
Data in use  Loaded in memory and being processed Trusted execution environments / secure enclaves, memory protection, tokenisation; emerging homomorphic encryption

EXAM ALERT: In transit → TLS/VPN. 

At rest → encryption plus access control. 

In use → enclaves and minimised exposure. 

Data in use is the hardest to protect because it must be readable to be processed. 

8. Data remanence and secure destruction

This is the most heavily tested cluster in the domain; learn it precisely. Data remanence is the residual data left on media after you think you’ve removed it. Deleting a file or running a quick format clears the pointer, not the data itself, so it stays recoverable. ‘Delete’ does not mean ‘gone’. 

The authoritative model is NIST SP 800-88, which defines three levels of sanitisation by how hard recovery becomes: 

Level  What it does  Media reusable?
Clear  Overwrite with new data using standard write commands; defeats simple recovery. A single pass suffices on modern drives. Yes
Purge  Make recovery infeasible even with lab techniques: degaussing, cryptographic erase, or firmware secure-erase. Sometimes
Destroy  Physically ruin the media: shred, pulverise, incinerate, melt, disintegrate. The most thorough option. No

The traps live in matching the method to the media: 

  • Magnetic drives (HDD) and tape: Overwrite, degauss, or destroy. Degaussing works here because the data is held magnetically (it usually ruins the drive for reuse too).
  • Solid-state drives (SSD) and flash: Do not degauss: there’s no magnetic field to disrupt, so it achieves nothing. Use the drive’s cryptographic erase or ATA secure-erase, or physically destroy it. Wear-levelling also makes plain overwriting unreliable on SSDs. 
  • Paper: Cross-cut shred, pulp, or incinerate. 
  • Optical media (CD/DVD): Shred or pulverise; you can’t overwrite them. 
  • Crypto-shredding: If the data was encrypted, destroying the encryption key makes the ciphertext permanently unrecoverable. Fast and ideal for cloud and SSDs where physical destruction isn’t practical. 

One more: the old DoD 5220.22-M multi-pass overwrite is legacy. NIST SP 800-88 is the current standard, and a single overwrite pass clears modern drives; multiple passes are no longer required. 

EXAM ALERT: You cannot degauss an SSD. Degaussing only affects magnetic media. For SSDs and for encrypted data, use cryptographic erase or physical destruction. And ‘format’ or ‘delete’ never counts as sanitisation. 

MANAGER’S LENS: Match the method to three things: the data’s sensitivity, the media type, and whether the media will be reused, resold, or discarded. Destroy for the crown jewels and for anything leaving your control; clearing is fine for reuse inside the same trust boundary.

9. Asset retention: end-of-life and end-of-support

Assets age out, and two dates matter: 

  • End-of-life (EOL): The vendor stops selling or producing the product. 
  • End-of-support (EOS, or end-of-service-life): The vendor stops issuing patches. This is the security cliff, because newly discovered vulnerabilities will never be fixed. 

Running EOS systems means consciously accepting unpatchable risk. Track EOL/EOS dates in the asset inventory, plan refresh cycles ahead of them, and isolate or compensate for anything you genuinely can’t replace in time. 

EXAM ALERT: End-of-support is the security-relevant date, not end-of-life. An unsupported but ‘still working’ system is a growing liability, not a cost saving. The exam frames this as a risk-acceptance decision for management.

10. Determining data security controls: scoping and tailoring

You rarely invent controls from scratch. You take a baseline, a starting control set for a given sensitivity or system type, drawn from a framework, and adapt it to your environment through two distinct steps: 

  • Scoping: Reviewing the baseline and removing controls that don’t apply to you (no wireless controls if you run no wireless). 
  • Tailoring: Adjusting the controls that remain to fit your risk: strengthening them, substituting compensating controls, or setting their parameters. 

Alongside this, standards selection means choosing the right framework or benchmark for the context and your compliance obligations: ISO/IEC 27002, NIST SP 800-53, or the CIS Controls, for example. 

EXAM ALERT: Scoping removes what doesn’t apply; tailoring adjusts what’s left. Swapping the two definitions is a reliable distractor. 

11. Data protection technologies

Know what each control actually does, because the exam tests the boundaries between them.

  • DLP (Data Loss Prevention): Discovers sensitive data and detects or blocks its unauthorised movement at the network egress, on endpoints, and in storage. It leans on data being classified or identifiable to work well. 
  • DRM / IRM (Digital / Information Rights Management): Persistent, policy-based protection that travels with the file: encryption plus usage rules, so restrictions on viewing, printing, forwarding, or expiry still apply after the file leaves your network. 
  • CASB (Cloud Access Security Broker): A control point between users and cloud/SaaS apps that gives visibility, enforces policy, and extends DLP and encryption into the cloud. 

Then the privacy-oriented methods, which are easy marks if you keep them straight: 

Method  What it does  Reversible?
Encryption  Scrambles data; readable only with the key  Yes, with the key
Tokenisation  Swaps a sensitive value for a token; the real value sits in a secure vault Yes, via the vault
Masking  Hides part of a value (e.g., shows only the last four digits) for display or testing No (just hidden)
Pseudonymisation  Replaces identifiers with a reversible pseudonym; a separate key can re-identify Yes, still personal data
Anonymisation  Irreversibly strips identifiers so data can’t be linked to a person  No

EXAM ALERT: Anonymisation is irreversible (no key), so truly anonymous data falls outside most privacy law. Pseudonymisation is reversible (a key exists), so it stays regulated personal data under GDPR. Tokenisation relies on a vault lookup; masking merely hides characters. 

Quick revision: commonly confused pairs 

The distinctions the exam reaches for in this domain. If you can state each row in a sentence, the questions here become straightforward. 

Concept A  Concept B  The one-line difference
Clearing  Purging  Overwrite for reuse vs. make lab recovery infeasible (degauss / crypto-erase)
Purging  Destruction  Recovery infeasible, media may survive vs. media physically ruined
Degaussing  SSD sanitisation  Works on magnetic HDD/tape vs. useless on SSD; use crypto-erase or destroy
Delete / format  Sanitisation  Removes the pointer (recoverable) vs. removes the data
Data at rest  Data in transit  Encryption + access control vs. TLS / VPN
Data in transit  Data in use  TLS / VPN vs. secure enclaves / memory protection
Anonymisation  Pseudonymisation  Irreversible, no key vs. reversible with a key (still personal data)
Tokenisation  Encryption  Vault swap, no math on the value vs. reversible with a key / algorithm
Masking  Tokenisation  Hides characters for display vs. substitutes with a vaulted token
DLP  DRM  Stops data leaving the org vs. protection that travels with the file
DRM  CASB  Per-file usage control vs. visibility & control over cloud apps
Data owner  Data custodian  Classifies & is accountable vs. implements & maintains controls
Data owner  Data steward  Protection & classification decisions vs. data quality & meaning
Data subject  Data user  The person the data is about vs. someone who uses the data
Scoping  Tailoring  Remove controls that don’t apply vs. adjust the ones that remain
End-of-life  End-of-support  No longer sold vs. no longer patched (the security date)
Retention  Remanence  How long you keep data vs. residual data left after deletion
Legal hold  Retention schedule  Freeze deletion for litigation vs. routine keep-then-dispose
Single-pass (800-88)  Legacy multi-pass  One pass clears modern drives vs. old DoD 3-pass overwrite
Tangible asset  Intangible asset  Hardware / media you can touch vs. data, IP, licences, reputation

Numbers and standards to lock in 

  • NIST SP 800-88 is the sanitisation standard: Three levels: Clear, Purge, Destroy. A single overwrite pass clears modern drives. 
  • You cannot degauss an SSD: Degaussing affects magnetic media only. For SSDs and encrypted data: cryptographic erase or physical destruction. 
  • Three data states: At rest, in transit, in use; each needs its own controls. 
  • Commercial classes: Public, Sensitive, Private, Confidential. Government: Unclassified, SBU/CUI, Confidential, Secret, Top Secret. 
  • Anonymisation = irreversible (no longer personal data). Pseudonymisation = reversible with a key (still personal data under GDPR). 
  • Owner classifies; custodian protects: Classify to the most sensitive element present. 
  • Crypto-shredding = destroy the key to render encrypted data unrecoverable. 
  • EOL = no longer sold; EOS = no longer patched (the security cliff).

Frequently Asked Questions

How much of the CISSP exam is Domain 2?

Around 10%. It is one of the lighter domains by weight, but dense with precise, factual distinctions that are easy to score if you've drilled them.

What's the difference between clearing, purging, and destroying data?

They're the three sanitisation levels in NIST SP 800-88. Clearing overwrites data so it survives simple recovery but the media is reusable. Purging makes recovery infeasible even in a lab (degaussing, cryptographic erase). Destruction physically ruins the media so it can never be used again.

Can you degauss an SSD?

No. Degaussing disrupts a magnetic field, and SSDs store data in flash cells, not magnetically, so it does nothing. To sanitise an SSD, use its cryptographic erase / secure-erase function or physically destroy it.

What is data remanence?

It's the residual data left behind on storage media after a delete or format. Because deletion typically removes only the file pointer, the underlying data remains and can be recovered; which is why proper sanitisation matters.

What's the difference between anonymisation and pseudonymisation?

Anonymisation irreversibly removes identifiers so data can no longer be linked to a person, and truly anonymised data falls outside most privacy laws. Pseudonymisation replaces identifiers with a reversible token or key, so re-identification is possible, meaning it's still regulated personal data under GDPR.

Who is responsible for classifying data?

The data owner, a business role accountable for the information. The data custodian then implements the controls that match the assigned classification.

What are the three states of data, and how do you protect each?

At rest (stored), encryption plus access control. In transit (moving over a network), TLS or a VPN. In use (in memory), secure enclaves, memory protection, and minimised exposure. Data in use is the hardest because it must be readable to be processed.

What is crypto-shredding?

Encrypting data and then destroying the encryption key, which makes the remaining ciphertext permanently unrecoverable. It's a practical way to 'destroy' data on SSDs and in the cloud, where physical destruction isn't feasible.

What's the difference between DLP, DRM, and CASB?

DLP stops sensitive data from leaving the organisation. DRM/IRM wraps protection and usage rules around a file so they persist wherever the file travels. CASB sits between users and cloud apps to give visibility and enforce policy, extending DLP and encryption into SaaS.

What's the difference between scoping and tailoring?

Scoping removes baseline controls that don't apply to your environment; tailoring adjusts the controls that remain to fit your specific risk. You scope first, then tailor.

TOP