EVOPEDIA

← all systems · see it in The System

Security orchestration, automation and response (SOAR)

A group of cybersecurity technologies that allow organizations to respond to some incidents automatically: it collects inputs monitored by the security operations team (alerts from the SIEM, threat intelligence platforms and other tools) and helps define, prioritize and drive standardized incident response activities (src:wp:security-orchestration). Researchers define security orchestration as 'the planning, integration, cooperation, and coordination of the activities of security tools and experts to produce and automate required actions in response to any security incident across multiple technology paradigms' (src:islam-2019-mlr). Walked as the platform and the practice: integrations (connectors, APIs, plugins), data processing and enrichment, case and ticket management with a knowledge base, the playbook engine that orchestrates tools and people, and the analysts who write the playbooks and approve what is not automated.

Properties published scienceour proposal

Sourced fields, and our own readings.

aliasesSOAR and automation
born at2015 or 2017 (term; C-3983); security orchestration literature from 2007 (src:islam-2019-mlr) and response-automation research from the early 2000s (src:huang-2025-intent-soar) as forerunners
born precisionyears
boundaryInside: the integration layer (connectors, wrappers, plugins, API gateway, tool registry), data processing and enrichment, the semantic layer and knowledge base, ticket and case management, the orchestrator and task manager that run playbooks (workflows, runbooks), the threat-intelligence management engine, dashboards and the user interface, and the playbooks themselves. Outside: the security tools it drives and reads (SIEM, IDS/IPS, EDR, firewalls, threat-intelligence platforms, ticketing systems, which other walks own), the organisation's estate, the attackers, and the SOC and incident responders who write and approve. Whether the analysts are part of SOAR or users of it is contested (C-3982): the review literature calls the human 'an essential part of a security orchestration platform'; the encyclopedia calls SOAR 'a group of cybersecurity technologies'.
nested levelA technological system (rung 12): software running inside a security operations centre (its response and coordination floors), inside an organisation; built on the APIs of the security tools it connects; its top floor is run by people.
scaletechnological

Measures (29) published science

alerts reaching the average SOC (Dark Reading, cited)10,000 alerts per day
false-positive rate, upper figures citedas high as 80 percent
survey of 500 decision-makers on alerts35% missed an alert; 22% ignored alerts entirely; 25% did not act on a high-profile alert percent of respondents
true positive alert ratio in one dataset (cited)1.59 percent
incident timeline that motivated the field: breach to discovery, then to remediation (BakerHostetler, cited)61 days to discover; 41 more days to remediate
alerts actively investigated (report cited)less than 10 percent
market forecast1.6 billion by 2021 (forecast) USD
Navy SOAR user study size24 participants; 6 commercial tools; over 200 hours people; tools; hours
customization per tool at installation2-4 hours per tool
tools completing a playbook with internet denied1 of 6 tools
ticket accuracy against the manual baselineall six tools worse than the baseline comparison
APIs to accommodate per tool when integrating (cited)over 200 APIs per tool
analysts' comfort with AI agents, response phase against identification phase2.71 against 4.04 (p < 0.001) mean rating
share of incident response operations automatable to some extent (vendor claim, cited)80-90 percent
work automated by one vendor (vendor claim, cited)800,000 man-hours in two years; $38.5 million hours; USD
attacker engagement time, static against SOAR-placed honeypots102 against 3148 seconds per instance
CACAO playbooks generated from CSAF advisories71 of 79; precision 98.72%; recall 78.57% playbooks; percent
AI-for-SOC papers focused on detection65 (of 189 papers); 88% non-explainable approaches percent
honeypot orchestration over four days of live traffic40 orchestrations; 7823 attacks; 965 DDoS packets; 3 malicious samples counts over 4 days
bandwidth while monitoring at rest0.373 ± 0.621 MB per 1,000 events
bandwidth to work one investigation22.80 ± 15.73 MB
LLM use in one SOC over 10 months3,090 queries from 45 analysts queries; analysts
API recommendation for building integrations (APIRO)91.9 percent Top-1 accuracy
Gartner forecast of adoption by 2019 (cited)30 percent of large and medium enterprises (forecast)
malware triage time in the Navy test range, unassisted against SOAR-assistedabout 15 minutes against almost instantaneous
vendor configuration window in the Navy test13 business days; hands-on only the final 7
'insufficient automation level of SOC components' raised, analysts against managers33.33 against 77.78 percent of participants
mitigation time targets by priority, and success rate, in one internal SOC (P11)P0 30 min, P1 60 min, P2 120 min, P3 24 h, P4 48 h; 87% met minutes; hours; percent
the SOC clocks a SOAR's cases feed (named, values site-specific)mean time to detect, react, triage, escalate, contain and resolve; number of incidents handled automatically named clocks

Open contentions (5) open contention

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

Cross-scale patterns seen here (16) our proposal

Shapes that recur when the same kind of problem is solved at different scales, and how each shows in Security orchestration, automation and response (SOAR). Each links to every other system that shows it.

Valence: what it trades, and with whom (13) 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 (11)

Sits inside / holds (3)

expressed by (25)

floor filled by (17)

Kinds (10)

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.

SOAR as security operations engineers and analysts describe it: the estate's alerts and the tools that raise them, up to the response decisions it carries out (reader's view)

  1. L1: The organisation's estate and what happens in it: hosts, networks, users and the attacks on them, as alerts and records
  2. L2: The security tools it drives and reads: SIEM, IDS/IPS, EDR, firewalls, threat-intelligence platforms, ticketing and team communication
  3. L3: Integration: connectors, wrappers, plugins, API gateway and tool registry
  4. L4: Data processing: alerts filtered, parsed and enriched with context and threat intelligence
  5. L5: Cases and knowledge: ticket and case management, the knowledge base of tools, capabilities and plans
  6. L6: Orchestration: the playbook engine (orchestrator and task manager) chooses and runs the workflow across tools
  7. L7: The analysts and the organisation's policy: they write the playbooks, approve what is not automated, and decide; the user interface is where they meet it

the boardroom (shared floors)

  1. L5: Cases and knowledge: ticket and case management (the incident-response platform function), the knowledge base of tools, capabilities and plans, and the record of past incidents
  2. L6: Orchestration: the playbook engine maps an alert to a playbook (by rules the analysts wrote), runs its steps across tools, and holds or forwards steps that need approval
  3. L7: The analysts and the organisation's policy: what is automated and to what level, which actions need a person, playbook design and approval, the decision on unknown threats

way in: Alerts from the SIEM and other detectors

  1. L1: Activity in the estate that a detector flags
  2. L2: The detectors: the SIEM, IDS/IPS, endpoint tools
  3. L3: Ingested through the integration (connector or API) for that tool
  4. L4: Filtered for syntactic and semantic correctness, then parsed

way in: Analysts' commands, approvals and overrides

  1. L1: The analyst's judgement, entered by hand at the console
  2. L2: The user interface and its APIs

way in: Answers to its own queries (enrichment and investigation results from tools)

  1. L1: The state of hosts, addresses and accounts in the estate
  2. L2: The tools queried: endpoint defence, scanners, reputation services
  3. L3: Queries and replies through the API gateway and wrappers
  4. L4: Context extracted and generated around the alert

way in: Attacker behaviour on its decoys (honeypot telemetry)

  1. L1: An attacker's moves inside the internal network
  2. L2: Honeypots it deployed record the activity, botnet traffic and DDoS

way in: Its own state and metrics on the dashboard

  1. L1: Its scenarios, assets and runs, counted
  2. L2: Shown on the dashboard
  3. L5: Case dispositions (false positive, true positive, benign true positive) and timestamps: the raw data of the SOC's measures

way in: Model predictions (ML trained on past incidents)

  1. L1: The record of past incidents

way in: Playbooks and standard procedures written for it

  1. L1: The organisation's incident response plans, vendors' advisories and shared playbooks
  2. L2: Written as text, flow diagrams, tables, or machine-readable markup (CACAO)
  3. L3: Generated or imported from structured advisories (CSAF to CACAO)
  4. L4: Interpreted against the knowledge base so playbooks from different vendors share one meaning
  5. L5: Stored in the knowledge base as IRP activities

way in: Results of the actions it ran (success, stall, error)

  1. L1: The estate after an action: blocked, isolated, or unchanged
  2. L2: The tool that executed the command answers (or does not)
  3. L3: Reply or error returned through the integration

way in: Status of the connected tools (the tool registry)

  1. L1: The tools available in the organisation and their capabilities
  2. L3: The tool registry discovers, registers and watches them

way in: The organisation's policy and rules on what may run without a person

  1. L1: Organisational policies and rules

way in: Threat intelligence feeds

  1. L1: Other organisations' and researchers' observations of threats, published as indicators
  2. L2: Feeds and threat-intelligence platforms
  3. L3: Taken in through feed integrations
  4. L4: Aggregated and deduplicated, then matched against alerts

way out: Containment: block addresses, isolate hosts, change firewall rules

  1. L1: Traffic blocked or a device cut off on the network
  2. L2: The firewall or endpoint tool carries it out
  3. L3: Sent through the integration to the tool's API
  4. L6: A playbook step (automatic or approved) commands the action

way out: Decoys deployed: honeypots placed where the attacker moves

  1. L1: A false target appears on the network
  2. L6: The engine decides to deploy a honeypot based on the attacker's behaviour

way out: Notify and assign: alerts, approval requests and assignments to analysts and stakeholders

  1. L1: Light from screens to the analyst's eyes
  2. L2: Shown on the dashboard or sent through team communication tools
  3. L6: Notification, assignment or a request for approval

way out: Removal: a phishing email or malware taken out, each by its own tool

  1. L1: The message or file removed from the mail store or host
  2. L6: An IRP step 'remove phishing email' or 'remove malware'

way out: Threat intelligence pushed downstream to the SIEM and firewall

  1. L1: Indicators installed in the receiving tools
  2. L6: Aggregated intelligence distributed

way out: Tickets and case documentation

  1. L1: The written ticket (in the tool or in Jira-like ticketing)
  2. L5: Case fields auto-populated from the investigation

Sources (23) published science

Information security operations center (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01)
Islam, C., Babar, M. A. & Nepal, S. (2019). A Multi-Vocal Review of Security Orchestration. ACM Computing Surveys 52(2). doi:10.1145/3305268 (arXiv:2002.09190; read in full through the paper portal)
Vielberth, M., Böhm, F., Fichtinger, I. & Pernul, G. (2020). Security Operations Center: A Systematic Study and Open Challenges. IEEE Access 8:227756-227779. doi:10.1109/ACCESS.2020.3045514 (CC BY 4.0; University of Regensburg repository copy) (behind its links)
Bridges, R. A., Rice, A. E., Oesch, S., Nichols, J. A. et al. (2023). Testing SOAR tools in use. Computers & Security 129:103201. doi:10.1016/j.cose.2023.103201 (sections read: introduction, experimental design 3.1, results 6.1-6.6) (behind its links)
Tilbury, J. & Flowerday, S. (2024). Humans and Automation: Augmenting Security Operation Centers. Journal of Cybersecurity and Privacy 4(3):388-409. doi:10.3390/jcp4030020 (CC BY) (behind its links)
Bartwal, U., Mukhopadhyay, S., Negi, R. & Shukla, S. (2022). Security Orchestration, Automation, and Response Engine for Deployment of Behavioural Honeypots. IEEE DSC 2022. doi:10.1109/DSC54232.2022.9888808; arXiv:2201.05326 (sections read: abstract, introduction, results) (behind its links)
Kokulu, F. B., Soneji, A., Bao, T., Shoshitaishvili, Y., Zhao, Z., Doupé, A. & Ahn, G.-J. (2019). Matched and Mismatched SOCs: A Qualitative Study on Security Operations Center Issues. ACM CCS 2019. doi:10.1145/3319535.3354239 (author copy) (behind its links)
Security orchestration (redirect from Security orchestration, automation and response) (Wikipedia, via Wikimedia Enterprise snapshot 2026-11-01) (behind its links)
Horneman, A. & Ray, J. (2021, 1 March). Benefits and Challenges of SOAR Platforms. Carnegie Mellon University Software Engineering Institute blog (institutional research blog; read in full) (behind its links)
Shaked, A., Cherdantseva, Y., Burnap, P. & Maynard, P. (2023). Operations-informed incident response playbooks. Computers & Security 134:103454. doi:10.1016/j.cose.2023.103454 (CC BY; sections read: abstract, introduction, background 2.A) (behind its links)
Huang, Z., Robin, J., Herbaut, N., Ben Rabah, N. & Le Grand, B. (2025). Toward an Intent-Based and Ontology-Driven Autonomic Security Response in Security Orchestration Automation and Response. arXiv:2507.12061 (sections read: abstract, introduction) (behind its links)
Lekidis, A., Mavroeidis, V. & Fysarakis, K. (2024). Towards Incident Response Orchestration and Automation for the Advanced Metering Infrastructure. arXiv:2403.06907 (behind its links)
Forsberg, J. & Frantti, T. (2023). Technical performance metrics of a security operations center. Computers & Security 135:103529. doi:10.1016/j.cose.2023.103529 (sections read: the case-SOC description, metrics table discussion, the passages on automation bias in metrics and free-text notes) (behind its links)
Empl, P., Schlette, D., Stöger, L. & Pernul, G. (2023). Generating ICS vulnerability playbooks with open standards. International Journal of Information Security. doi:10.1007/s10207-023-00760-5 (CC BY; sections read: abstract, CACAO background, evaluation, limitations) (behind its links)
Giarimpampa, D., Meier, R., Bissyandé, T. F. et al. (2025). Exploring the Role of Artificial Intelligence in Enhancing Security Operations: A Systematic Review. ACM Computing Surveys. doi:10.1145/3747587 (behind its links)
Singh, R., Tariq, S., Jalalvand, F., Baruwal Chhetri, M., Nepal, S. et al. (2025). LLMs in the SOC: An Empirical Study of Human-AI Collaboration in Security Operations Centres. arXiv:2508.18947 (sections read: abstract, introduction) (behind its links)
Sworna, Z. T., Islam, C. & Babar, M. A. (2023). APIRO: A Framework for Automated Security Tools API Recommendation. ACM Transactions on Software Engineering and Methodology 32(1). doi:10.1145/3512768 (behind its links)
Islam, C. (2020). Architecture-centric support for security orchestration and automation. PhD thesis, University of Adelaide (sections read: chapter 1, chapter 2 sections 2.3-2.6, chapter 3 sections 3.1-3.3) (behind its links)
Kinyua, J. D. M. & Awuah, L. (2021). AI/ML in Security Orchestration, Automation and Response: Future Research Directions. Intelligent Automation & Soft Computing 28(2). doi:10.32604/iasc.2021.016240 (behind its links)
Kremer, R., Wudali, P. N., Momiyama, S., Araki, T., Furukawa, J., Elovici, Y. & Shabtai, A. (2023). IC-SECURE: Intelligent System for Assisting Security Experts in Generating Playbooks for Automated Incident Response. arXiv:2311.03825 (sections read: abstract, introduction) (behind its links)
Srinivas, S., Kirk, B., Zendejas, J., Espino, M., Boskovich, M., Bari, A., Dajani, K. & Alzahrani, N. (2025). AI-Augmented SOC: A Survey of LLMs and Agents for Security Automation. Journal of Cybersecurity and Privacy 5(4):95. doi:10.3390/jcp5040095 (CC BY; sections read: abstract, capability-maturity model, integration challenges) (behind its links)
González-Granadillo, G., González-Zarzosa, S. & Diaz, R. (2021). Security Information and Event Management (SIEM): Analysis, Trends, and Usage in Critical Infrastructures. Sensors 21(14):4759. doi:10.3390/s21144759 (PMC8309804, CC BY) (behind its links)
Shaikh, A. J., Bastian, N. D. & Shah, A. (2026). A Red Teaming Framework for Evaluating Robustness of AI-enabled Security Orchestration, Automation, and Response. arXiv:2605.17075 (behind its links)