EVOPEDIA

← all systems · see it in The System

Trust boundary

A boundary where program data or execution changes its level of trust, or where two principals with different capabilities exchange data or commands; inside it all sub-systems have equal trust (src:wp:trust-boundary). In security engineering the line is drawn around what must be trusted (the trusted computing base, whose bounds 'equate to the security perimeter', src:tcsec-1985) and is enforced by a guard at a door in a wall (src:saltzer-schroeder-1975). Walked as the line, the checks made at it and the practice that draws it, from kernel and process boundaries to browser origins, network perimeters and zero trust's per-request checks; borders and checkpoints on the shelf.

Properties published scienceour proposal

Sourced fields, and our own readings.

born atby 1972
born precisionabout 5 years (rings 1971-72; reference monitor 1972)
boundaryInside: the line between parts trusted to different degrees and everything that holds it: the wall that forces every reference through a door, the door itself (gates, designated entry points, interfaces, policy enforcement points), the guard's checks at it (authentication, validation, mediation), the trusted side's mechanisms (the trusted computing base, the reference validation mechanism, the policy engine) and the practice that draws the line (threat models with trust boundaries marked). Outside: the parts on either side (users, programs, networks, devices, data stores), what crosses (data and control), intruders, the access-control decision rules (walked as sys:access-control), the network devices that build one kind of perimeter (sys:firewall), and the wider security policy and mission that set what is to be protected.
nested levelA design element inside computer systems (rung 12): in hardware (privilege modes and rings), the operating system (kernel and process boundaries), applications and browsers (sandboxes, origins, site isolation), networks and enterprises (perimeters, micro-perimeters); drawn by people in threat models (rung 11 practice). Part of computer security and information security; the place where access control is applied.
scaletechnological

Measures (13) published science

characteristic clock: one user-to-kernel crossing (getpid)1000-1500 cycles, of which about 100 are the switch itself (70 in, 40 back); about 150 cycles on the L3 microkernel CPU cycles
size of the seL4 kernel (ARMv6)10.2 (9.7 C, 0.5 assembler); original L4 6.4; Fiasco.OC 37.6 thousand lines of code
one-way IPC across a protection boundary, seL4 on ARM11188 CPU cycles
verification limit on a kernel's sizeabout 9,000 source lines of code ('pushed the limits')
new site-isolation bugs from one month of fuzzing4 (one with full control of the victim site; $8000 bounty) bugs
site-isolation bypasses by cause, Chrome and Firefox39 known (text); table rows: 28 missing checks, 4 bypassed checks, 6 origin confusion = 38 vulnerabilities
rings: Multics against today's systemsMultics 8; x86 4; most operating systems use 2 (ring 0 and 3) privilege rings
normal and alerting access counts in NIST's contextual example20 to 30 records a day normal; alert above 100 records per day
work factor of four-letter passwords26^4 = 456,976 tries
element types in a threat-modelling data flow diagram5 (four DFD types plus trust boundaries) element types
readings of 'trust boundary' in DFDs3 (trust or privilege levels; attacker-model assumptions; deployment) interpretations
added cost of user-kernel crossings with KPTI0.28% (KAISER's authors); about 5% for most workloads, up to 30% (a Linux developer); PostgreSQL read-only 7-17% (16-23% without PCID) percent overhead
organisations planning zero trust capabilities in 202072% share of organisations

Open contentions (6) open contention

Where sources or readings disagree. Both stay on the map.

Cross-scale patterns seen here (20) our proposal

Shapes that recur when the same kind of problem is solved at different scales, and how each shows in Trust boundary. Each links to every other system that shows it.

Valence: what it trades, and with whom (9) published science

Each exchange with a neighbour, read from both sides: who gains, who pays, and what carries it.

Partners and parts published science

Built from / builds (9)

Sits inside / holds (3)

expressed by (22)

floor filled by (22)

Kinds (13)

Notes published science

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.

the boardroom (shared floors)

  1. L5: Mediation and enforcement at the door: the reference validation mechanism, the guard, the policy enforcement point; tamper proof, always invoked, small enough to analyse
  2. L6: What is trusted and by what rule: the trusted computing base (its bounds are the security perimeter), the security policy, the policy engine and its trust algorithm
  3. L7: The protection goal: the authority structure the owners want carried out, the information and resources protected, the mission

trust boundaries as security engineers draw them: parts of a system trusted to different degrees and what passes between them, up to the protection goals the boundary serves (reader's view)

  1. L1: Parts of a system with different trust: users and programs, processes, devices, networks, data stores, external entities; inside a zone all have equal trust
  2. L2: The line: where the level of trust changes (a ring boundary, a process wall, an origin, a network perimeter, a drawn boundary on a data flow diagram)
  3. L3: What crosses: data from an untrusted source, control with a change of privilege, commands between principals
  4. L4: The checks at the door: authentication, validation, the authority check
  5. L5: Mediation that cannot be bypassed: the reference validation mechanism, the enforcement point, with the trust zone behind it kept small
  6. L6: What must be trusted and the rule it applies: the trusted computing base and the security policy; in zero trust the policy engine
  7. L7: The goal: the authority structure the owners want, information protected from unauthorized use

way in: Code and updates arriving from the build side (dependencies, compilers, drivers, vendor updates)

  1. L1: Code written by others
  2. L2: Built by compilers, assemblers and loaders that are themselves code
  3. L3: Signed: author and integrity bound to the artifact
  4. L4: Verified when deployed: the author's or build system's identity and a checksum

way in: Context: time, place, the subject's history, threat intelligence

  1. L1: Time and geolocation of the request
  2. L2: Recorded: the policy database, past behaviour, threat intelligence
  3. L3: Kept as state on each subject by the policy engine
  4. L4: Compared with the subject's normal pattern

way in: Control passing in: calls from a less-privileged ring, process or domain

  1. L1: A program in an outer ring or user mode asks for something only an inner ring may do
  2. L2: A system call or cross-ring call: the processor's mode bit and the gate
  3. L3: The crossing itself checked by hardware; calls land only at designated entry points
  4. L4: The inner side's entry point: the protected subsystem's procedure

way in: Data arriving from the less-trusted side (user input, network sockets, files, partner feeds)

  1. L1: Data from an untrusted source: user input, a network socket, an internal API or stored record
  2. L2: Received at the component's interface: the place the boundary is drawn on the data flow diagram
  3. L3: Parsed by a maintained parser for the expected format
  4. L4: Validated for syntax and semantics against an allowlist, wherever it crosses

way in: People and goods arriving at a checkpoint (borders; shelf)

  1. L1: Travellers and freight moving across borders
  2. L2: Met by officers and machines at the checkpoint (officers ride Human's sight channel)
  3. L4: Sorted and filtered by criteria, slowed or sped up; biometric passports read at automated gates

way in: Signs of compromise on the far side (malformed messages, failed checks)

  1. L1: A compromised renderer can send arbitrary messages
  2. L2: Received by the privileged browser process over IPC
  3. L3: Checked against the site the browser process tracks for that renderer
  4. L4: Judged malformed or invalid: a sign of compromise

way in: The designers' model of the system and its threats (where the line is drawn)

  1. L1: What the team knows of the system: 'what are we working on?'
  2. L2: Drawn as a data flow diagram: external entities, processes, data stores, flows
  3. L3: Trust boundaries added as the fifth element
  4. L4: STRIDE prompts applied to components, flows and boundaries
  5. L5: Mitigations chosen and placed (the readings of the boundary differ, so placement can go wrong)

way in: The state of the asking device (posture)

  1. L1: Enterprise-owned and personal devices, patched or not
  2. L2: Monitored for integrity and security posture
  3. L3: The CDM system reports the asset's state to the policy engine
  4. L4: Weighed into the confidence level for this request

way in: Where web content came from (its origin)

  1. L1: Content created by many authors, some malicious, fetched from servers by URI
  2. L2: Tagged with its URI as it is fetched
  3. L3: Grouped into origins: same scheme, host and port
  4. L4: Run in the sandboxed renderer process for its site; the privileged browser process checks every message

way in: Who is asking: the principal and its credentials

  1. L1: A person or program on the far side; the principal is who is accountable
  2. L2: A credential presented (password, device artifact)
  3. L3: Over a path untrusted software cannot imitate (trusted path)
  4. L4: Authenticated and authorized for this request

way out: Admitted across: entered at a controlled point with least privilege

  1. L1: The resource used; beyond the PEP is the trust zone hosting it
  2. L3: Control enters the domain at its entry point
  3. L5: Granted for this request only

way out: Contained: the compromised side cut off or killed

  1. L1: The compromise held on its side
  2. L3: Connection terminated by the PEP
  3. L5: Compromise recognised

way out: Recorded: logs and audit

  1. L1: Rejected input recorded without secrets
  2. L6: Changes to the decision rules logged

way out: Redrawn: the line moved, shrunk or split (design change)

  1. L1: The built system: rings, processes, perimeters, micro-perimeters
  2. L3: Functions moved across rings (drivers out of the kernel; vDSO into processes; some logic into the kernel for speed)
  3. L6: Shrink the trust zone, shrink the TCB
  4. L7: The goal: protect the resources that matter most

way out: Refused: default is no access

  1. L1: Values outside the rules rejected
  2. L5: No permission found: refused

way out: Waved through or slowed (checkpoints; shelf)

  1. L1: Some movement accelerated and waved through, some slowed, diverted or halted
  2. L5: Sorted into categories

where trust boundaries sit, from the hardware out to the organisation (placement reading, defence in depth)

  1. L1: Hardware and firmware: the processor's most privileged level and its secure areas
  2. L2: Hypervisor
  3. L3: Kernel: the user-kernel line crossed by system calls
  4. L4: Process: sandboxes and per-site renderer processes
  5. L5: Application and origin
  6. L6: Network: perimeters and micro-perimeters
  7. L7: Organisation: compartments and need-to-know

Sources (41) published science

Trust boundary (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01)
Drescher J, Klein DA, Johns M (2026) Are your sites truly isolated? Automatically detecting logic bugs in site isolation implementations. NDSS 2026 (doi:10.14722/ndss.2026.240902); sections read: abstract, I-III (behind its links)
Rose S, Borchert O, Mitchell S, Connelly S (2020) Zero Trust Architecture. NIST Special Publication 800-207 (doi:10.6028/NIST.SP.800-207); sections read: 1-3, 5, 7, appendix B (behind its links)
Trusted computing base (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Protection ring (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Leese M, Wittendorp S (2018) The new mobilities paradigm and critical security studies: exploring common ground. Mobilities 13(2) (doi:10.1080/17450101.2018.1427016); sections read: introduction and the three overlaps (behind its links)
Saltzer and Schroeder, The Protection of Information in Computer Systems, Proceedings of the IEEE 63(9), 1975 (doi:10.1109/PROC.1975.9939) (behind its links)
OWASP Cheat Sheet Series: Threat Modeling Cheat Sheet (CC BY-SA 4.0), GitHub source as of 2026-11-22 (behind its links)
Elphinstone K, Heiser G (2013) From L3 to seL4: what have we learnt in 20 years of L4 microkernels? SOSP '13 (doi:10.1145/2517349.2522720); sections read: 1-3.1, 3.2, 4 (IPC fast path), tables 1-2 (behind its links)
Sion L, Yskout K, Van Landuyt D, van den Berghe A, Joosen W (2020) Security threat modeling: are data flow diagrams enough? ICSEW'20 (doi:10.1145/3387940.3392221) (behind its links)
Kernel page-table isolation (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Campbell M (2020) Beyond zero trust: trust is a vulnerability. Computer 53(10) (doi:10.1109/MC.2020.3011081); column, text as reached (13 kB) (behind its links)
Anderson JP (1972) Computer Security Technology Planning Study, Vol. I (ESD-TR-73-51) (behind its links)
US Department of Defense (1985). Trusted Computer System Evaluation Criteria (DoD 5200.28-STD, the Orange Book) (behind its links)
Schroeder MD, Saltzer JH (1972) A hardware architecture for implementing protection rings. Communications of the ACM 15(3) (doi:10.1145/361268.361275); full text via the paper portal; sections read: abstract, the abstract description of rings, privileged instructions, the layered Multics supervisor. (Draft cited the ACM SIGOPS OSR copy, doi:10.1145/850614.850623, abstract only; SOSP '71 is doi:10.1145/800212.806498 per the specialist, not looked up) (behind its links)
OWASP Cheat Sheet Series: Input Validation Cheat Sheet (CC BY-SA 4.0), GitHub source as of 2026-11-22 (behind its links)
Reference monitor (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Barth A (2011) The Web Origin Concept. RFC 6454, IETF (doi:10.17487/RFC6454); sections read: 1-3 (behind its links)
Sandbox (computer security) (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Joint Task Force (2020) Security and Privacy Controls for Information Systems and Organizations, NIST SP 800-53 Rev. 5 (doi:10.6028/NIST.SP.800-53r5); sections read: AC-8 System Use Notification (behind its links)
Thompson 1984, Reflections on Trusting Trust, Communications of the ACM 27(8):761-763, doi:10.1145/358198.358210 (behind its links)
Code signing (Wikipedia) (behind its links)
Trusted execution environment (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Schroeder MD (1972; thesis page says Ph.D. 1973) Cooperation of mutually suspicious subsystems in a computer utility. MIT EE PhD thesis (MAC TR-104), DSpace@MIT via OpenAlex (W1561721839); sections read: chapter 1 (objectives, protected subsystems, mutually suspicious subsystems) (behind its links)
Capability Hardware Enhanced RISC Instructions (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Same-origin policy (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Teerakanok S, Uehara T, Inomata A (2021) Migrating to zero trust architecture: reviews and challenges. Security and Communication Networks (doi:10.1155/2021/9947347); sections read: 1-2 and the BeyondCorp review (behind its links)
Ghasemshirazi S, Shirvani G, Alipour MA (2023) Zero Trust: applications, challenges, and opportunities. arXiv:2309.03582 (preprint); sections read: 1 and challenges (behind its links)
Threat model (Wikipedia, via Wikimedia Enterprise snapshot) (behind its links)
Forrest, Hofmeyr and Somayaji, Computer Immunology, Communications of the ACM 40(10), 1997 (doi:10.1145/262793.262811) (behind its links)
Syed NF, Shah SW, Shaghaghi A, Anwar A, Baig Z, Doss R (2022) Zero Trust Architecture (ZTA): a comprehensive survey. IEEE Access (doi:10.1109/ACCESS.2022.3174679); sections read: abstract and introduction (behind its links)
Zero trust architecture (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Biba Integrity Model (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Bell–LaPadula model (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Hardy N (1988) The confused deputy (or why capabilities might have been invented). ACM SIGOPS OSR 22(4) (doi:10.1145/54289.871709); author's open copy at cap-lore.com, read in full (behind its links)
Confused deputy problem (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Time-of-check to time-of-use (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Lampson BW (1973) A note on the confinement problem. Communications of the ACM 16(10) (doi:10.1145/362375.362389); text via the paper portal (11 kB, the whole note as far as can be seen) (behind its links)
Covert channel (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Greshake K, Abdelnabi S, Mishra S, Endres C, Holz T, Fritz M (2023) Not what you've signed up for: compromising real-world LLM-integrated applications with indirect prompt injection. arXiv:2302.12173 (preprint); sections read: abstract and introduction (defender-side framing only) (behind its links)
Honeypot (computing) (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)