Blog / Deep Research Report

ENISA vs. TLCTC: A Systematic Gap Analysis & Integration Opportunity

How the Top Level Cyber Threat Clusters (TLCTC) framework addresses the structural definitions gaps in ENISA's current cybersecurity approach.

TR
TLCTC Research Team
• • Updated • Loading read time...
Executive Summary

This analysis examines how the European Union Agency for Cybersecurity (ENISA) addresses cyber risk and cyber threat definitions, identifying structural gaps in their current framework that the Top Level Cyber Threat Clusters (TLCTC) framework could address. ENISA operates with a hybrid classification approach that is predominantly outcome-based rather than cause-oriented, creating significant conceptual conflation between threats, techniques, and impacts. The TLCTC framework, with its cause-oriented, actor-agnostic taxonomy based on 10 non-overlapping generic vulnerabilities, directly addresses this foundational deficit.

Updated November 3, 2026

This post was checked against ENISA’s own documents and TLCTC v2.7. Corrected: the 2016 taxonomy has eight high-level threats, not nine. The prime-threat structure described in Part I was used up to ETL 2024; ETL 2025 replaced it. Four quotations that could not be traced to an ENISA source were removed or rephrased, and the risk-framework claim attributed to ENISA’s 2022 Compendium was withdrawn. The cluster table now quotes each generic vulnerability verbatim, and the ransomware example uses v2.7 notation. The argument is unchanged. For the 2026 edition, see ENISA Threat Landscape 2026: Counting Incidents Without a Cause.

Solving the structural semantic problem in ENISA's threat classification.
Conceptual Model: Integration of TLCTC with ENISA Cyber Framework.

Part I: ENISA's Current Approach to Cyber Threats and Risk

1.1 ENISA's Foundational Threat Definition

ENISA’s 2016 Threat Taxonomy scopes cyber threats as threats applying to assets related to information and communication technology. Most of them materialise in cyberspace; some materialise in the physical world but affect information and cyber assets. ENISA’s current threat-landscape methodology adds one operative definition, in a footnote. A cyberattack is “any intentional effort to disrupt, compromise, or damage digital assets, systems, or networks through unauthorized access or malicious actions” (CTL Methodology, September 2026, p. 6). The definition names an intent and its effects. It does not name the weakness exploited.

This broad scoping creates immediate boundary challenges—physical sabotage of network infrastructure sits alongside software vulnerabilities in the same framework, without clear differentiation of the generic vulnerability being exploited.

1.2 The Eight-Category ENISA Threat Taxonomy

The ENISA Threat Taxonomy was first established in 2016 and, according to ENISA’s methodology, updated in 2022. The spreadsheet ENISA serves on its threat-taxonomy page organises threats into eight high-level categories, with some 175 threats beneath them:

# Category Examples
1Physical attack (deliberate/intentional)Fraud, sabotage, vandalism, theft, terrorist attacks
2Unintentional damage / loss of information or IT assetsInformation leakage due to human error, erroneous use or administration of devices and systems
3Disaster (natural, environmental)Earthquakes, floods, fires, dangerous radiation leaks
4Failures/MalfunctionSoftware and hardware failures, supplier failure
5OutagesNetwork outages, outages of mobile networks, cooling outages
6Eavesdropping/Interception/HijackingInterception of information, interception of compromising emissions
7Nefarious Activity/AbuseIdentity theft, denial of service, malicious code, social engineering, manipulation of information, targeted attacks (APTs), data breaches
8LegalViolation of laws or regulations, failure to meet contractual requirements, abuse of personal data

The categories do not share a single axis. Spear phishing attacks appears twice, under Social Engineering and under Targeted attacks (APTs etc.), and so does Mobile malware. Web application attacks / injection attacks (SQL, XSS) are filed under Malicious code/software/activity, and Compromising confidential information (data breaches), an outcome, is listed as a threat.

1.3 The ETL Prime Threats (up to 2024): Outcome-Oriented Classification

Up to ETL 2024, the annual ENISA Threat Landscape (ETL) reports grouped incidents into “prime threats” for strategic communication. ETL 2024 lists the seven below. Its text speaks of “eight prime threat types” and, a page later, of “the seven areas mentioned above”. This structure reveals ENISA’s predominantly outcome-based approach:

Prime Threat Classification Nature (Problem)
RansomwareAttack type + business model + outcome conflated
MalwareTool category (overlaps with ransomware)
Social EngineeringAttack methodology spanning multiple stages
Threats Against DataOutcome category (data breach/leak) — not a cause
Threats Against AvailabilityAttack technique + outcome conflated
Information ManipulationCampaign type + outcome
Supply Chain AttacksVector category (closest to cause-based)
Double-Counting and Statistical Skew

ETL 2024 states that incidents involving more than one threat category “were thus analysed in the context of all respective categories.” A ransomware attack delivered via phishing that exfiltrates data could therefore count under ransomware, malware, social engineering and threats against data. ETL 2025 replaced the prime threats with a “revised and concise format” built on incident types, intrusion vectors and actor groups, and ETL 2026 kept it. The mixing of axes remained. A second skew comes from volume: of the 4,875 incidents in ETL 2025, DDoS made up 76.7%, almost all claimed by hacktivists. The number of low-impact claims, not their severity, sets the ranking.

1.4 ENISA's Risk Framework Approach

ENISA’s risk work follows standard formulations: Risk = Likelihood × Impact, extended in the ISO 27005 tradition to combine impact, threat likelihood and vulnerability level. In these formulas the threat is an input to likelihood. Nothing in them keeps the cause of an event apart from its consequences, which is the separation a bow-tie makes explicit.

1.5 ENISA's Acknowledged 'Language Problem'

A 2015 publication by ENISA and the Cyber Security Coordination Group (CSCG), titled ‘Definition of Cybersecurity – Gaps and overlaps in standardisation’, directly confronts the terminological chaos:

"Finding a common understanding is a major challenge and it might not be possible to harmonize the definition and usage of the term."

The paper notes that even the spelling (‘Cybersecurity’ or ‘cyber security’) depends on the organisation. The threat-landscape reports carry caveats of their own: ETL 2025 states that 28.5% of its incidents could not be associated with a specific sector.

Part II: Identified Gaps in ENISA's Current Framework

2.1 Structural Conflation Problems

The most significant gap is mixing conceptual levels without clear methodological justification. Their threat categories, from the taxonomy to the ETL headline categories, conflate:

  • Root causes (vulnerability exploited) with attack mechanisms (technique used).
  • Tools (malware) with specific malware types (ransomware listed separately).
  • Techniques (social engineering) with outcomes (threats against data).
  • Actor motivations with threat categories: ETL 2026 organises its deep-dive chapters by actor type (cybercrime, State-nexus, FIMI, hacktivists) and then reports that “the distinction between threat categories continued to blur.”

2.2 Absence of Cause-Based Classification

ENISA’s methodology calls its taxonomy “the guiding principle for CTLs and the standard for threats referenced within ENISA.” Its high-level threats mix causes, events and consequences rather than following one classification principle. The framework currently lacks:

  • Explicit generic vulnerability identification for each threat category.
  • Non-overlapping classification criteria (mutual exclusivity).
  • Attack path sequencing capability (attack steps as ordered sequences).
  • Velocity analysis for control effectiveness assessment.
  • Domain boundary crossing notation.

2.3 No Native Bow-Tie Integration

ENISA's ISO 27005-aligned approach treats cause-consequence distinction as implicit rather than explicit. This creates challenges for control mapping (preventive vs. mitigating), risk register development, and business impact analysis that doesn't conflate threat type with outcome type.

2.4 Incomplete Harmonization Despite Significant Investment

ENISA’s harmonization efforts remain structurally incomplete. The Reference Security Incident Taxonomy (RSIT), which the methodology cites for the 2022 update, is a separate classification for incident exchange between CSIRTs, built on the earlier eCSIRT.net taxonomy. That leaves two taxonomies side by side rather than one converged vocabulary. Significantly, the CTL methodology (August 2025, updated September 2026) states that the ENISA threat taxonomy “is currently under revision for the purpose of developing a more mature, actionable framework.”

Part III: How TLCTC Addresses ENISA's Gaps

3.1 TLCTC's Cause-Oriented Architecture

The Top Level Cyber Threat Clusters framework provides exactly what ENISA's current approach lacks: a cause-oriented, actor-agnostic taxonomy that classifies threats by the generic vulnerability initially exploited.

# TLCTC Cluster Generic Vulnerability (verbatim, v2.7)
1Abuse of FunctionsThe inherent trust, scope, and complexity designed into software functionality and configuration.
2Exploiting ServerServer-side implementation flaws enable unintended behavior.
3Exploiting ClientClient-side implementation flaws enable unintended behavior.
4Identity TheftInsufficient binding, at the point of authentication, between a presented credential and the authentic holder of the identity it claims.
5Man in the MiddleThe lack of sufficient control over the communication path.
6Flooding AttackFinite capacity limitations inherent in any system component.
7MalwareThe software environment’s designed capability to execute potentially untrusted foreign code.
8Physical AttackPhysical accessibility of infrastructure and the exploitability of physical-layer properties.
9Social EngineeringHumans can be influenced into unsafe actions or decisions.
10Supply Chain AttackTrust in third-party components and update channels can be subverted.

Clusters #1–#7 are internal to the software domain; #8, #9 and #10 are bridge clusters that cross from the physical, human and third-party domains. Credential acquisition is classified in whichever cluster enabled it; #4 is the use of a credential to authenticate (R-CRED).

3.2 Key Differentiators: TLCTC vs. ENISA

Dimension ENISA Approach TLCTC Approach
Classification BasisOutcome-oriented (ransomware, data breach)Cause-oriented (generic vulnerability exploited)
Mutual ExclusivityOverlapping categories (double-counting)Non-overlapping (one step = one cluster)
Attack SequencesNot natively supportedAttack paths, e.g. #9 → #4 → #1 → #7 + [DRE: Ac]
Velocity AnalysisNot addressedVC-1 to VC-4 classes with Δt notation
Bow-Tie IntegrationNot presentCore architectural anchor
Strategic/OperationalSeparate frameworks (ETL + ATT&CK)Unified dual-layer architecture

3.3 The Ransomware Example: Classification Comparison

Consider a typical ransomware incident: phishing email → credential theft → lateral movement → data encryption.

ENISA Classification

  • Problem: Under the ETL 2024 prime threats, analysed under Ransomware, Malware, Social Engineering, AND Threats Against Data.
  • Result: Double-counting of the same incident across multiple categories.
  • Missing: No attack sequence visibility.

TLCTC Classification

#9 ||[human][@External→@Org]|| + [DRE: C] → #4 → #1 → #7 + [DRE: Ac]
  • Precision: Each step classified by generic vulnerability exploited. The phishing step discloses the credential; using it is #4 (R-CRED).
  • Clarity: “Ransomware” names the outcome: data present but unusable, [DRE: Ac]. The cause at that step is #7, the encryptor executing (R-EXEC).
  • Velocity: #9 →[Δt=5m] #4 →[Δt=2h] #1 →[Δt=15m] #7 (illustrative)

Part IV: How ENISA Would Benefit from TLCTC Integration

4.1 Solving the 'Language Problem'

TLCTC provides the semantic foundation ENISA needs by establishing 10 non-overlapping threat categories based on generic vulnerabilities. This eliminates the classification ambiguity of the current taxonomy and its roughly 175 threats, and separates causes from outcomes via Bow-Tie architecture.

4.2 Enhancing the ETL Methodology

ENISA's ETL reports could integrate TLCTC as a complementary classification layer:

1. ETL Incident Types and Actor Chapters
Strategic Communication (Current outcome- and actor-oriented approach)
2. TLCTC Clusters
Causal Analysis (New layer, cause-oriented)
3. MITRE ATT&CK
Operational Detail (Current TTP-oriented approach)

4.3 Improving Risk Assessment Consistency

ENISA's ISO 27005-aligned risk formula (Risk = Impact × Threat Likelihood × Vulnerability Level) would gain precision through TLCTC integration:

  • Threat Likelihood: Assessed per TLCTC cluster rather than per outcome category.
  • Vulnerability Level: Mapped to generic vulnerabilities (#1-#10) rather than specific CVEs alone.
  • Control Effectiveness: Evaluated using attack velocity classes (VC-1 to VC-4).

4.4 Supporting NIS2 and DORA Compliance

The EU's NIS2 Directive and DORA regulation require consistent threat and risk reporting. TLCTC enables standardized incident classification across EU member states, comparable risk registers using cause-based taxonomy, consistent control mapping via TLCTC-to-NIST-CSF integration, and attack path documentation for critical infrastructure sectors.

4.5 Addressing ENISA's 'Under Revision' Taxonomy

Given that ENISA's taxonomy is currently under revision, TLCTC offers a ready-made solution that has been systematically designed from axioms. It includes axiomatic foundations for consistent interpretation, provides a JSON schema for machine-readable incident documentation, maps all 706 MITRE ATT&CK Enterprise v20 techniques at the operational layer, and is available under CC BY 4.0 license, aligning with ENISA's open collaboration model.

Conclusion: A Strategic Opportunity

ENISA's cybersecurity frameworks represent the most comprehensive European effort to standardize threat and risk terminology. However, fundamental tensions remain between operational utility and conceptual precision. TLCTC is not proposed as a replacement for ENISA's frameworks, but as a complementary layer that provides the cause-oriented precision currently missing. The integration would strengthen ENISA's position as the EU's cybersecurity authority while addressing the fundamental taxonomic gaps that limit cross-border incident comparability and risk assessment consistency.

ENISA Gap TLCTC Solution
Outcome-based classification creating double-countingCause-based classification with mutual exclusivity
No attack sequence capabilityAttack path notation (#9→#4→#1→#7)
Absent Bow-Tie integrationBow-Tie as core architectural anchor
Implicit domain boundariesExplicit Domain Boundary Operators
Acknowledged 'language problem'Semantic clarity through axiomatic definitions

Recommended Next Steps

  • Engage ENISA's Threat Landscape team during their taxonomy revision process.
  • Propose a TLCTC mapping of ETL incident types and intrusion vectors as pilot integration.
  • Offer TLCTC JSON schema for STIX 2.1 integration compatibility.
  • Demonstrate attack path notation value for NIS2/DORA compliance reporting.

References

  1. ENISA Threat Landscape 2026 (September 2026)
  2. ENISA Threat Landscape 2025 (October 2025; v1.3, September 2026)
  3. ENISA Threat Landscape 2024 (September 2024)
  4. ENISA Cybersecurity Threat Landscape Methodology (August 2025, updated September 2026)
  5. ENISA Threat Taxonomy (2016; spreadsheet retrieved November 2026)
  6. ENISA / CSCG, Definition of Cybersecurity – Gaps and overlaps in standardisation (December 2015)
  7. Reference Security Incident Taxonomy (RSIT) Task Force
  8. TLCTC v2.7 core paper (the December 2025 original was written against the v2.0 white paper)