NewCTI-CRAFT: advanced threat intelligence training for regulated testing.CTI-CRAFT trainingRegister interest
ThreatInsights Plan a test

Financial sector

One in five banking cyber incidents originate from third parties

The ECB puts the figure at around 20%. What it does and does not tell you, and what it means for TIBER-EU and DORA TLPT scenarios.

APAndy PlayfordFounder and principal analyst5 min read
Bottom line

Approximately one in five reported cyber incidents affecting euro-area banks originates from third-party providers, according to the European Central Bank. That is a warning about dependency and concentration risk — not evidence that one in five incidents is a malicious supplier compromise. Financial institutions need to understand both the services they depend on and the adversaries capable of exploiting those dependencies.

The European Central Bank's November 2026 analysis, Decoding the cyber-AI nexus, highlights a significant exposure within Europe's financial system: roughly 20% of reported cyber incidents affecting euro-area banks originate from third-party service providers.

The figure matters, but its meaning is easily distorted. A supplier-origin incident may involve a technology failure, misconfiguration or service disruption. It is not automatically an intrusion, nor does it establish an adversary's capability to exploit the provider.

For financial institutions, the more useful question is: which external dependencies support our critical business functions, and what would happen if one of them failed or was compromised?

Third-party concentration creates shared exposure

Banks increasingly depend on external cloud platforms, payment processors, software suppliers, managed service providers and specialist operational technology. The same provider may support many institutions, creating a shared point of exposure.

Concentration matters particularly where services are difficult to replace, recovery is complex or a supplier has privileged access to an institution's systems. Strong internal security controls cannot by themselves prevent disruption to an external service on which the institution relies.

01IdentifyMap suppliers supporting critical financial functions.
02EvaluateAssess concentration, privileges and substitutability.
03DistinguishSeparate operational failure from malicious activity.
04CorrelateLink relevant adversary capability to actual exposure.
05ValidateChoose resilience exercises or evidence-led TLPT scenarios.

A supplier inventory is a starting point, not an intelligence assessment. Institutions need to understand which dependencies are critical, how they interact and what consequences could follow from their disruption.

Not every cyber incident is a cyberattack

One of the most important analytical distinctions is between an ICT or cyber incident and a malicious cyberattack. Operational faults and non-malicious failures can have serious financial-sector consequences without an attacker being involved.

Conversely, compromise of a supplier identity, trusted management platform or privileged integration can create a credible adversarial path into an institution.

Both cases deserve investigation, but they demand different responses. A service outage may be best assessed through resilience, recovery and supplier-continuity exercises. A technically supported adversary path may warrant threat-led testing.

Conflating the two produces weak intelligence: it exaggerates adversary activity, obscures the real cause of incidents and risks generating scenarios that are not supported by evidence.

What financial-sector threat intelligence should ask

Traditional reporting often catalogues threat actors, malware, vulnerabilities and campaigns. For regulated financial institutions, that information is most valuable when connected to their operating environment and critical business functions.

  • Critical dependencies: Which suppliers and services support payments, trading, settlement, customer access or other important functions?
  • Trust relationships: Which third parties have administrative privileges, persistent connectivity or authority over sensitive workflows?
  • Adversary relevance: Which actors have demonstrated capability against comparable supplier environments?
  • Attack prerequisites: What access, conditions and technical weaknesses would need to exist for a proposed path to work?
  • Consequences: Could the outcome affect availability, integrity, confidentiality or the delivery of a critical financial service?

The difference between an interesting incident and useful intelligence is the ability to show why it matters to a particular institution, with source confidence and assumptions made explicit.

What this means for TIBER-EU and DORA TLPT

Threat-led penetration testing is intended to examine realistic adversary behaviour against critical functions. TIBER-EU and DORA TLPT make intelligence-led scenario development central to that process.

Imagine a hypothetical bank relying on an externally managed payment platform. An intelligence assessment may establish that the platform supports a critical service, a third party maintains privileged access and a relevant adversary has demonstrated methods for abusing similar trusted relationships.

That combination could justify deeper analysis of a potential scenario. But dependency alone is not proof of an exploitable attack path. The intelligence team must assess whether the adversary capability is evidenced, whether the prerequisites apply to the institution and whether the proposed activity is within the agreed test scope.

Some supplier risks belong in operational resilience or tabletop exercises rather than TLPT. Selecting the right validation method is part of sound analytical judgement.

AI adds another dimension, but not an automatic explanation

The ECB also examines the intersection of AI and cyber risk. AI may alter the speed and scale of some attacker activities, while financial institutions' own use of externally hosted models and AI services introduces additional dependencies.

Neither observation proves AI was involved in a specific incident. Intelligence assessments should identify actual capabilities, access conditions and evidence before attributing activity to AI-enabled adversaries.

Key judgements
  1. Third-party incident origins are a material financial-sector exposure, but the ECB's approximately 20% figure must not be presented as a measure of malicious supplier compromise. High confidence
  2. Supplier concentration can create correlated disruption across multiple institutions, particularly where critical services are difficult to substitute. High confidence
  3. Connecting institution-specific dependencies to documented adversary capabilities will improve scenario prioritisation; a plausible path still requires evidence before being used for regulated testing. Moderate confidence
So what Choose one critical financial service and map the external providers, privileged relationships and recovery dependencies behind it. Then ask two separate questions: what happens if the service fails, and what evidence supports a realistic adversary route through it? The answers should determine whether the next step is resilience validation, threat intelligence collection or a threat-led test.

What ThreatInsights would assess

In an intelligence-led engagement, the useful output is not a list of suppliers or a generic warning about third-party risk. It is a defensible assessment of the organisation's exposure to relevant adversaries and the potential consequences.

  • Priority intelligence requirements: Define which critical functions and supplier dependencies need assessment.
  • Source evaluation: Separate documented adversary activity from unverified claims and analytical hypotheses.
  • Scenario prerequisites: Identify the permissions, access paths and environmental conditions required.
  • Confidence and alternatives: Record uncertainty, competing explanations and gaps in the evidence.
  • Validation handover: Specify whether findings support TLPT scenario development, a tabletop exercise or further supplier assurance.

For managed CTI, the same discipline supports ongoing monitoring of material changes in adversary capability and relevant exposure. For TIBER-EU and DORA TLPT, it supports realistic scenarios grounded in the institution's critical functions.

Conclusion

The ECB's findings should prompt financial institutions to examine the limits of their visibility beyond their own infrastructure. Strong internal controls are necessary, but they do not eliminate dependencies on external providers.

Yet the answer is not to label every supplier failure an attack. Effective intelligence distinguishes operational disruption from malicious activity, connects documented threats to actual exposure and makes the evidence behind each judgement visible.

Knowing your own attack surface is essential. Understanding the wider ecosystem on which your critical services depend is equally important.

Sources and assessment notes

  • Primary source: European Central Bank, Decoding the cyber-AI nexus, 16 November 2026.
  • Methodology context: ECB TIBER-EU framework. Established framework, not a new November TLPT requirement.
  • Interpretation: The approximate 20% statistic concerns reported incident origins, not verified malicious supplier compromises. The payment-platform example is hypothetical and does not claim a real incident or verified attack chain.
AP
Andy PlayfordFounder and principal analyst, ThreatInsights. Leads threat intelligence for TIBER-EU, DORA TLPT and iCAST engagements.LinkedIn

Planning a TIBER-EU or DORA TLPT test?

Talk to a practising threat intelligence lead. Thirty minutes, no pitch deck, and a straight answer on scope and timing.