Risk Management • Methodology • Feb 2026

Software Failure through the Bow-Tie Lens

BK
Bernhard Kreinz
TLCTC Framework v2.0

The TLCTC framework defines a generic risk management methodology — a Bow-Tie structure with cause-based classification, Data Risk Events, and NIST CSF function verbs — that is not limited to adversarial threats.

Adversarial

Cyber Risk Event

TLCTC Threat Clusters #1–#10. Attacker intent drives the causal chain exploiting generic vulnerabilities.

Non-Adversarial

IT Risk Event

Software Failure (INT + 3PC). No attacker—the inherent imperfection of software is the cause.

The Bow-Tie — Software Failure

← Cause Side · Preventive
Central Event
Consequence Side · Mitigating →
INT

Internal Software Failure

  • • Code defects & logic errors
  • • Misconfigurations
  • • Resource leaks
3PC

Third-Party Software Failure

  • • COTS product defects
  • • SaaS provider outage
  • • Library / OSS regressions
Shared Generic Vulnerability: Inherent imperfection of software.
Pivot Point Software Failure
LoI
Integrity
LoAc
Accessibility
LoAv
Availability
LoC
Confidentiality
Business Impact
  • → Revenue loss
  • → SLA breach penalties
  • → Trust erosion

NIST CSF Mapping

Verbs describing what controls do. Click a function to analyze the control architecture (examples only).

INT — Internal (Local)
3PC — Third-Party (Local)
Umbrella — Enterprise

Adversarial · TLCTC #10

Supply Chain Attack

An adversary intentionally compromises a trusted third-party component. The generic vulnerability is the organizational dependency on that trust.

||[@Vendor→@Org]|| #10→#7→#1→[LoC+LoI]

Non-Adversarial · SW Failure

Third-Party Software Failure

No adversary. The trusted component fails due to its own latent defects. The structural problem is identical: you accepted code you cannot fix.

||[@Vendor→@Org]|| SW Failure→Service Disruption

The Generic Risk Management Method

The TLCTC framework establishes a generic risk management methodology whose structural components — the Bow-Tie model, cause-based classification, and Data Risk Events — are domain-independent.

When applying this method to "Software Failure", the sub-type differentiation (INT vs. 3PC) becomes the critical innovation. A single control objective demands structurally different strategies depending on whether you own the code or accepted it through trust.

Process Flow