← all systems · see it in The System
Information security
Information security (infosec) is the practice of protecting information by mitigating information risks; it is part of information risk management (src:wp:information-security). NIST defines it as 'the protection of information and systems from unauthorized access, use, disclosure, disruption, modification, or destruction in order to provide confidentiality, integrity, and availability' (src:nist-sp800-12r1). Protected information may take any form: electronic or physical, tangible (paperwork) or intangible (knowledge). Walked here as the practice (its goals, people, standards, laws, models and history) together with what it protects; the computer-side reading of the same practice is sys:computer-security (walked wave 13), whose floors this walk points at instead of repeating.
Properties published scienceour proposal
Sourced fields, and our own readings.
| born at | contested (C-1357): 'since the early days of communication'; or WWII and the computer era |
| born precision | none; no dated first case in the sources read |
| boundary | Inside: the goals (confidentiality, integrity, availability, and the attributes and definitions argued over them), the work of protecting information (identifying and classifying assets, assessing risk, choosing and running controls of physical, technical and administrative kinds, monitoring, incident handling, business continuity, change management), the people who do it (owners, risk executives, CISOs, auditors, security staff, every employee who handles information) and the bodies of standards, models and law that shape it. Outside, protected by it: the information itself in any form (paper, electronic, knowledge, genomes) and the organisations, states and people whose mission it serves; the adversaries and accidents it guards against (walked from the attacker's side in computer-security's secondary stack); the tools it borrows (cryptography, operating systems, networks). The edge with computer security and cyber security is contested (C-1156): 'cyber security goes beyond the boundaries of traditional information security' (src:von-solms-2013), while the encyclopedia files computer security as its subdiscipline. |
| nested level | A practice carried by people and organisations (rung 11), since the 1960s largely run on computers and networks (rung 12); part of information risk management; serves organisations' missions; shaped by law, markets and states. Built from controls, roles, standards and its subdiscipline computer security; borrows cryptography as a tool. |
| scale | social |
Measures (14) published science
| characteristic clock: crowd-sourced phishing reports | about 10% within 5 min; 20% within 15 min; 30-40% within 30 min share of all reports of a simulated phish, by time after delivery |
| employees who clicked at least one simulated phish (15 months) | 4,729 of 14,733 (32.10%) employees |
| simulated phishing emails clicked | 5.67 % of 117,864 sent |
| reports by employees in 15 months | 14,401 (11,035 of them simulated) emails reported |
| Gordon-Loeb bound on optimal security investment | < 1/e (about 37%) of the expected loss |
| Willemson counterexample to the GL bound | up to 50%; close to 100% when assumptions relax of the expected loss |
| FIPS 199 impact levels | 3 (LOW, MODERATE, HIGH) plus NOT APPLICABLE for confidentiality levels per security objective |
| security objectives in US federal law | 3 objectives (confidentiality, integrity, availability) |
| attributes in the Parkerian Hexad | 6 attributes |
| NIST control families | 20 information security and privacy control families (SP 800-53, as introduced in SP 800-12r1) |
| CISSP holders worldwide | 191,036 ISC2 members holding the CISSP (December 2025) |
| information-security behaviour studies in health information systems | 35 studies; 20 distinct theories studies (systematic review) |
| dimensions of information security culture | 7 dimensions (Roer and Petric 2017) |
| Anderson 1972 planning study duration | 2 February to 1 September 1972 study period |
Open contentions (6) open contention
Where sources or readings disagree. Both stay on the map.
- C-1118 In cybersecurity, do attackers have the advantage?
booked: The map keeps both: attack is easier because defenders must cover everything, and attackers have no inherent advantage because one block can break their chain.
the other reading: attacker advantage; no inherent advantage; defender advantage under intelligence-driven defence - C-1344 What does it mean for information to be secure?
booked: The map uses confidentiality, integrity and availability as standards and law state them, and keeps the wider six-part version and the 'appropriate access' definition alongside.
the other reading: CIA is the definition of information security (ISO, 44 U.S.C.); CIA is too broad and too narrow; security is each agent having 'just the appropriate access' relative to a stakeholder (Lundgren and Moller 2019); six atomic attributes are needed (Parker 1998), whose merits are debated - C-1348 Does training people right after they click a fake phishing email help?
booked: The map records the large field study that found it made employees more likely to click again, and keeps the earlier, smaller studies that found it helped alongside.
the other reading: embedded training improves resilience (earlier, smaller and shorter studies the paper cites as contradicted) - C-1354 How much should an organisation spend on protecting information?
booked: The map records the classic result, under about 37% of the expected loss, for the cases its model covers, and keeps the counterexamples showing it can be more alongside.
the other reading: the 1/e rule holds generally (Baryshnikov 2012, as reviewed); counterexamples: Hausken 2006; Willemson 2006 (up to 50%, close to 100% when assumptions relax) - C-1357 When did information security begin?
booked: The map dates the practice from protecting letters and messages, with its computer-era form much later, and keeps the reading that computer security started it alongside.
the other reading: it begins 'since the early days of communication' (marking, couriers, strong boxes); 'The establishment of computer security inaugurated the history of information security', with the need appearing in WWII (the same article, citing Whitman and Mattord) - C-1383 Are users the weakest link in security?
booked: The map keeps both: people as the most vulnerable point, and people rationally ignoring advice that costs them more than it saves.
the other reading: 'the most vulnerable point in most information systems is the human user' (the encyclopedia, 'research has shown'); users' rejection of advice is 'entirely rational from an economic perspective' (Herley 2009); users as a working sensor network (Lain et al.)
Cross-scale patterns seen here (21) our proposal
Shapes that recur when the same kind of problem is solved at different scales, and how each shows in Information security. Each links to every other system that shows it.
- barrier-gates-complexity (structural)
A closed environment (cleared users, protected consoles and circuits) let 1970 technology run a secure system; without that barrier, 'Contemporary technology cannot provide a secure system in an open environment', and Ware advised against putting classified information in such systems at all. The wi - nested ladder (structural)
Classification levels are a ladder of implication: a clearance at a level dominates the levels below (Bell-LaPadula's 'dominates' relation; government labels Unclassified to Top Secret; FIPS 199 LOW < MODERATE < HIGH). Half-rung: need-to-know cuts across the ladder (two top-secret holders still need - outside-to-inside (structural)
The partition moves from the surroundings into the system. Before: security was isolation, 'simply removing the entire system to a physical environment in which penetrability is acceptably minimized', a barrier round the machine room and its uniformly cleared users. After: open, multilevel use force - same role, different builds (structural)
One job, proving who you are, done by unrelated structures: something one knows (a password), has (a card) or is (fingerprints, retina). Classification does the same job with different label sets in business, government, cross-sector (TLP) and personal settings. - trade one property for another (structural)
Security is bought with productivity and availability: 'balanced protection ... without hampering organization productivity'; the CIA properties 'can even conflict'; time-locks buy security by withholding availability. Kinds settle at different points: FIPS 199 public information {C: NA, I: MODERATE - who opens the door (structural)
Access is a port built ahead of time by the receiver: identification, authentication and authorization configured from policy ('The access control mechanisms are then configured to enforce these policies'), and a reference monitor that validates every reference. Counter-case, open field: passive int - borrowed-then-held (partial)
The same move read as regulation: protection first supplied by the surroundings (buildings, guards, the national clearance apparatus: 'adequate standards and techniques exist and are used by the cognizant authority to insure the reliability of those cleared'), later held by the system's own tamper-p - causal ceiling (partial)
No one can see all the risk: 'It is not possible to identify all risks, nor is it possible to eliminate all risk'; breaches 'are generally rare and emerge in a specific context which may not be easily duplicated'; Saltzer and Schroeder: no complete method exists, tied to 'the negative quality of the - holding / entrainment (partial)
The organisation sets when its parts' access dies: a terminated employee's access is 'immediately disabled'; classifications are reviewed periodically; clearances are issued and withdrawn as jobs change (Ware). - lightning (partial)
A breach is a stored difference released at once: confidential information that 'fall[s] into the hands of a competitor or hacker' causes 'widespread, irreparable financial loss'; captured U-570 papers became an intelligence coup. After it the system is not the same. - namespace completeness (the map predicts its gaps) (partial)
Control catalogues and checklists (ISO 27002's eleven areas, NIST's 20 families) are read to find the controls an organisation is missing; the hexad claims every breach falls in one of six atomic attributes. - red-queen (partial)
New threats and vulnerabilities 'emerge every day', so risk management 'must be repeated indefinitely'; phishing 'showed no sign of slowing down' over two decades of research. Both sides improve; the stalemate is implied rather than measured (C-1118). - same shape, opposite sign (partial)
Withholding access from a party is the one act both ways: done to unauthorised parties it is the protection (access control, time-locks), done to authorised users it is the harm (denial of use, availability lost). - symmetry breaking (name the cut) (partial)
Before the *-property, information could flow either way across levels ('a malicious program ... might pass classified information along by putting it into an information container labeled at a lower level'); the *-property cuts the symmetry so information flows only up (or level), and everything la - templating (partial)
The CIA triad is a module copied unchanged across ISO standards, US federal law and other national standards, and controls catalogues (ISO 27002, NIST's 20 families) are reused by any organisation; national schemes (Cyber Essentials, Essential Eight) restate it. - The carry (partial)
Face 1, overflow passed on: risk the organisation will not hold is passed to the next holder ('the risk can be transferred to another business by buying insurance or outsourcing'), and insurance plus investment beats either alone below a premium threshold. Face 3, a lookahead that skips the wait: th - The Payment to Entropy (partial)
Each layer of protection pays a toll that never fully converts into security: investment buys security 'at a decreasing rate' (GL), indirect costs 'may exceed the direct cost of the control' (NIST), users pay in effort (Herley). Anderson names the thermodynamic form for defenders: 'Paddy has thermod - the whole sets the local reference (partial)
FIPS 199's high water mark: a system's category is the highest impact among the information it holds, and that level then sets the protection for all of it; the computer room is 'secured to the level commensurate with the most highly classified and sensitive material handled by the system' (Ware). - Umwelt: each system carves its own slice of one shared medium (partial)
The practice's channels carve its slice of the world: assets, threats, weaknesses, logs, people's reports, losses, infrastructure events, law, standards, culture, mission and its own models. Different sectors carve the same information with different labels (business, government, TLP). - conservation law plus closure (surface)
The Gordon-Loeb balance (maximise expected net benefit of investment) leaves the breach-probability function open; it predicts a bound only once a class of security breach function is chosen, and other closures (Hausken, Willemson) give other answers. - same toolkit, independent builds (surface)
Many national and sector bodies build their own schemes from the same parts (controls, CIA, risk): NIST CSF, UK Cyber Essentials, Australia's Essential Eight, PCI DSS, IEC 62443.
Valence: what it trades, and with whom (34) published science
Each exchange with a neighbour, read from both sides: who gains, who pays, and what carries it.
- Electrical grid via sys:computer-network · Information security gain
- Operating system via sys:electromagnetic-field · Information security gain
- Programming language via sys:written-record · Information security mixed
- Malware via sys:written-record · Information security gain
- Authentication via sys:written-record · Information security gain
- Access control via sys:written-record · Information security gain
- Intrusion detection system via sys:written-record · Information security gain
- Incident response via sys:written-record · Information security gain
- Threat actor via sys:computer-network · Information security mixed
- Security policy and governance via sys:organization · Information security gain
- DNA and genes via sys:sequential-information · Information security mixed
- Human (Homo sapiens) via sys:human-society · Information security mixed
- Government via sys:written-record · Information security gain
- Written record (durable external marks: texts, archives, canon) via sys:written-record · Information security gain
- Law (rules and law) via sys:written-record · Information security mixed
- Internet via sys:electromagnetic-field · Information security mixed
- Computer via sys:electromagnetic-field · Information security gain
- Market and trade via sys:human-society · Information security mixed
- Health care via sys:written-record · Information security mixed
- Cryptography via sys:written-record · Information security gain
- Computer security via sys:written-record · Information security gain
- Computer network via sys:electromagnetic-field · Information security mixed
- Risk management via sys:organization · Information security gain
- Insider threat via sys:organization · Information security gain
- Organization via sys:written-record · Information security gain
- Database via sys:written-record · Information security gain
- Cloud computing via sys:written-record · Information security mixed
- Software vulnerability via sys:written-record · Information security mixed
- Physical security via sys:written-record · Information security gain
- Privacy via sys:organization · Information security gain
- Secure software development via sys:software · Information security gain
- Cybercrime and its underground economy via sys:internet · Information security loss
- Business continuity via sys:organization · Information security gain
- Security awareness via sys:human-society · Information security mixed
Partners and parts published science
Built from / builds (15)
- Cryptography
- Computer security
- Access control
- Incident response
- Security controls (physical, technical, administrative)
- Information classification (labels and handling)
- Business continuity and disaster recovery
- Risk management
- Standards (ISO/IEC 27000 family, NIST FIPS and SPs, Common Criteria, PCI DSS)
- The profession (CISOs, auditors, certified practitioners)
- Formal security policy model
- Correspondents against state interceptors for the secrecy of letters
- The eternal war in memory (top-level competition)
- Campaign: control-flow hijacking against deployed protections
- Attackers against database defenders (SQL injection)
Sits inside / holds (19)
- Computer security
- Authentication
- Access control
- Intrusion detection system
- Security operations center (SOC)
- Incident response
- Security policy and governance
- Government
- Risk management
- Organization
- Insider threat
- Firewall (computing)
- Physical security
- Privacy
- Supply chain security
- Social engineering
- Penetration testing
- Security awareness
- Security audit and compliance
expressed by (26)
- Generation/Synthesis
- Templating / Standardization
- Internal Coordination/Regulation
- Sensing/Scanning
- Hypothesis/Goal Selection
- Mapping/Modeling
- Uncertainty Reduction/Learning
- Risk-managed Exploration
- Active Experimentation / Probing (Optimal Experimental Design)
- Trail-Marking / Externalized Memory (Stigmergy)
- Surprise / Change-Point Detection
- Frontier Outposts / Sensing Extension
- Social Learning / Public Information Use
- Reactive Response
- Proactive Defense
- Physical/Virtual Barrier
- Self-Monitoring
- Validation of Identity/Intention
- Validation of Effectiveness
- Enclave/Segregation
- Resilience & Repair
- Redundancy & Diversity
- Triage & Prioritized Allocation
- Deterrence
- Quality Control & Error Correction
- Bridging Supply Gaps (Store & Dormancy)
floor filled by (29)
- Operating system
- Computer security
- Access control
- Intrusion detection system
- DNA and genes
- Human (Homo sapiens)
- Government
- Written record (durable external marks: texts, archives, canon)
- Law (rules and law)
- Computer
- Market and trade
- Cryptography
- Security operations center (SOC)
- Incident response
- Security controls (physical, technical, administrative)
- Information classification (labels and handling)
- Business continuity and disaster recovery
- Risk management
- Standards (ISO/IEC 27000 family, NIST FIPS and SPs, Common Criteria, PCI DSS)
- The profession (CISOs, auditors, certified practitioners)
- Organization
- Security clearances
- Secrecy law (Official Secrets Acts)
- Formal security policy model
- Firewall (computing)
- Database
- Secure software development
- Supply chain security
- Security awareness
represents (1)
same practice as (1)
Kinds (4)
- Computer security
- Cyber security
- Information assurance
- Procedural security of paper records (marking, couriers, safes)
wired (1)
Notes published science
- In 1970 the US Defense Science Board's task force concluded that the technology of the day could not build a secure computer for 'an open environment' at all; the only safe computer was one walled off with cleared users in a guarded room. Two years later the Anderson study moved that wall inside the machine, as a 'reference monitor' checking every single access, the design idea still under today's operating systems (src:ware-1970, src:anderson-1972).
- In a 15-month study of 14,733 employees, a third clicked at least one fake phishing email, but the same people also worked as an alarm system: about 10% of their reports of a new phish came in within 5 minutes. The common industry fix, showing a training page to those who clicked, made them more likely to click again (src:lain-2022-phishing).
- The classic definition of information security (keep it secret, accurate and available) breaks on a bank's time-lock vault, which is secure precisely because nobody can get in until morning; two philosophers propose instead that information is secure when every person has 'just the appropriate access' relative to whoever has a stake in it (src:lundgren-moller-2019).
Layers
The ways it takes in its world, where they meet, and how it acts back, each from L1 (where it meets the world) up.
information security as practitioners describe it: information in any form up to the mission it serves (reader's view)
- L1: Information in any form (paperwork, electronic data, knowledge, genomes) and the world threats come from: people with intent, accidents, nature, failing infrastructure
- L2: Assets known: identified, valued, owned and classified (labels from Public to Top Secret, TLP colours)
- L3: Controls: physical (fences, guards, safes), technical (access control, encryption, the reference monitor) and administrative (policies, procedures, training); identification, authentication and authorization
- L4: Monitoring and incident handling: audit trails, continuous monitoring, user reports, incident response plans and teams
- L5: Risk management: categorise information by impact, assess threats and vulnerabilities, treat risk (accept, mitigate, transfer, deny), carry residual risk; business continuity
- L6: Policy, standards, law and culture; roles and responsibilities made explicit (risk executive, CIO, CISO, owners, users); formal models
- L7: Purpose: confidentiality, integrity and availability (or appropriate access for each stakeholder) in service of the organisation's mission
the boardroom (shared floors)
- L4: Monitoring and incident handling: continuous monitoring, audit review, report triage, incident response plans and teams (computer-security's b-detect and b-respond are this floor for computers)
- L5: Risk management: categorise, assess, choose controls, accept residual risk, plan continuity
- L6: Policy, standards, law, culture and explicit roles (risk executive, CIO, senior information security officer or CISO, owners, users); formal models adopted
- L7: Purpose and governance: the security goals (CIA, or appropriate access) in service of the mission, owned by senior management
the paper era: marked documents, trusted carriers and clearances (secondary stack)
- L1: Correspondence and documents carried between diplomats, commanders and offices
- L2: Markings on the document, trusted carriers, guarded strong boxes
- L3: Multi-tier classification and markings saying who may handle a document and where it is stored
- L4: Clearances, need-to-know and operating procedures (Ware's two organisational leakage points), proper destruction
- L5: Secrecy law and official classification systems (Official Secrets Act 1889; the Bell-LaPadula levels for computers)
way in: Audit trails and monitoring records
- L1: Every authentication and access attempt
- L2: An audit trail for every access
- L3: Continuous monitoring keeps awareness current
way in: Incidents, breaches and losses
- L1: A threat uses a vulnerability: harm, an impact
- L2: The impact read as a loss of confidentiality, integrity or availability
- L3: Counted and reported (ransomware in health care: disrupted care, integrity at risk; reporting limited)
- L4: The expected loss that investment is weighed against
way in: Information assets (what there is to protect)
- L1: Information in any form: paper, electronic, knowledge
- L2: Assets identified and valued: people, buildings, hardware, software, data, supplies
- L3: An owner named and a classification policy written
- L4: Each information type categorised by the impact of losing C, I or A
way in: Laws and regulation
- L1: Laws and regulations on how data is accessed, processed, stored, transferred and destroyed
- L2: Specific acts (Computer Security Act 1987, FISMA 2002 and 2014, HIPAA, GLBA, breach notification laws)
- L3: Liability: who carries it sets the care taken
- L4: Security objectives defined in law
- L5: Compliance requirements (breach notification, auditor attestation)
way in: People's attitudes, behaviour and culture
- L1: People perceive and decide about risk in different ways
- L2: What employees think and feel, and what they do
- L3: Individual and organisational factors measured
- L4: Seven dimensions of security culture (attitudes, behaviours, cognition, communication, compliance, norms, responsibilities)
- L5: Culture evaluated and changed in cycles
way in: People's reports of suspicious messages and events
- L1: Phishing and other suspicious emails reach employees in their working context
- L2: The employee judges it; type of computer use predicts vulnerability better than amount
- L3: A report button in the email client
way in: Physical and environmental events
- L1: Power failures, communications loss, water leaks, fire, flood, storms, strikes
- L2: System downtime, sometimes with the systems still working (staff can't reach them)
- L3: Physical protection, continuous in time
way in: Standards and frameworks
- L1: Published guidance against threats to a user's or organisation's cyber environment
- L2: Families of standards: ISO/IEC 27000 (ISMS), Common Criteria, NIST CSF, FIPS and SPs
- L3: Control catalogues: 20 control families
- L4: Certification of practitioners (CISSP)
- L5: Assessment criteria and certification of organisations
way in: The field's own models and research
- L1: Principles learned from experience, for want of a complete method
- L2: Formal models: the reference monitor; Bell-LaPadula's basic security theorem
- L3: Economic models: perverse incentives, the Gordon-Loeb investment model
- L4: Definitions argued (CIA, the hexad, appropriate access)
way in: The mission and its stakeholders' needs
- L1: The organisation's mission
- L2: The system's role in it
- L3: Who should have what access, relative to each stakeholder
- L4: Security requirements stated in mission terms
way in: Threat sources (adversarial and non-adversarial)
- L1: Acts of nature, war, accidents and malicious acts from inside or outside; passive interception of emanations
- L2: Threat sources named: malicious hackers and their subcategories, insiders, fraud, malicious code, errors, infrastructure loss
- L3: A threat assessment
- L4: The likelihood that a threat will use a vulnerability
way in: Vulnerabilities (weaknesses)
- L1: Weaknesses in people, process and technology
- L2: Errors, omissions and bugs, unnoticed in normal use
- L3: A vulnerability assessment, with the probability of exploitation
- L4: Controls themselves evaluated for vulnerabilities
way out: Awareness and training
- L1: Every employee and partner trained on the classification scheme (simulated phishing with embedded training: contested)
- L6: Regular security awareness training
way out: Breach notification and reporting
- L1: Reports filed (often late; reporting gives a limited view)
- L6: Notification required by law
way out: Changes made under change management
- L1: The change implemented at the appointed time
- L4: Each change tested before production, with a back-out plan
- L6: A change review board oversees changes
way out: Continuity and recovery
- L1: Critical ICT infrastructure recovered after a disaster
- L3: Backups scheduled, critical elements duplicated
- L5: Business continuity management
way out: Controls chosen and applied
- L1: On the medium: locks and safes, encryption of information in transit and in storage
- L3: Physical, technical and administrative controls in overlapping layers
- L5: Controls selected from the risk assessment
way out: Incident response: contain, revoke, recover
- L1: Access revoked: a terminated employee's access disabled and the person escorted off the premises
- L4: An incident response plan is activated
way out: Labels and handling rules written onto information
- L1: Marks on the information itself
- L2: Labels by sector: business, government, Traffic Light Protocol
- L6: A classification policy
way out: Risk transferred (insurance, outsourcing)
- L1: Cyber-insurance bought beside security investment
- L5: Leadership chooses to transfer a risk
Who fills each layer
- L3: Access control, Physical security
- L4: Intrusion detection system
- L5: Supply chain security
- L6: Security policy and governance
- L7: Computer security