Versatile AI Risk Assessment

AI threat catalogue Reliability and Responsible Use Production

Model Drift & Degradation

The quality of an AI model gradually declines in operation because the world keeps changing while the model stays frozen on old knowledge (model drift). Without dedicated monitoring, decisions get silently worse and nobody notices.

As of: October 2026 · Catalogue version 2026.10.24.2 · 7 mitigations · 9 verified sources

Description

A model learns from past data. As customer behaviour, language, products, or fraud patterns change, the learned relationships fit the present less and less. Specialists speak of distribution shift (the distribution of the input data moves) and concept drift (the learned relationship itself becomes outdated). Feedback loops can add to this: the model's outputs influence future input data and reinforce existing distortions. Where new models are increasingly trained on AI-generated content, quality can also decay across model generations (model collapse). Because the system keeps responding fluently and produces no error messages, the decline stays invisible for a long time without continuous monitoring.

Possible impact

Gradually degrading forecasts, scores, or filter decisions lead to lost revenue, poor planning, and undetected fraud, often over months. The EU AI Act obliges deployers of high-risk systems to monitor the system's operation, so unnoticed drift can also become a compliance issue. Fixing it requires retraining and reworking decisions that have already been made.

Example

A payment fraud detection model was trained on historical patterns. As fraudsters change their methods, the detection rate drops month by month; it only becomes apparent once the annual accounts show significantly higher losses.

Recommended mitigations (7)

Compare effects and implementation levels. Expand each entry for the additional classification and rationale.

  • Continuous performance monitoring

    Technical
    Effect
    Detective
    Implementation level
    Model & training, Use & operations
    Classification and rationale

    Complementary control typeOrganizational & process-based

    Primarily technical: Software or analytical tools systematically produce and evaluate measurements, deviations, or attack indicators; complemented by binding workflows.

  • Automated drift detection

    Technical
    Effect
    Detective
    Implementation level
    Model & training, Use & operations
    Classification and rationale

    Primarily technical: Software or analytical tools systematically produce and evaluate measurements, deviations, or attack indicators.

  • Periodic retraining schedules

    Organizational & process-based
    Effect
    Corrective
    Implementation level
    Model & training, Use & operations
    Classification and rationale

    Complementary control typeTechnical

    Primarily organizational and process-based: Defined selection, operating, or lifecycle procedures make the control binding and repeatable; complemented by technical implementation.

  • A/B testing for model updates

    Technical
    Effect
    Detective
    Implementation level
    Model & training, Use & operations
    Classification and rationale

    Primarily technical: Software or analytical tools systematically produce and evaluate measurements, deviations, or attack indicators.

  • Data quality monitoring

    Technical
    Effect
    Detective
    Implementation level
    Data, Model & training, Use & operations
    Classification and rationale

    Complementary control typeOrganizational & process-based

    Primarily technical: Software or analytical tools systematically produce and evaluate measurements, deviations, or attack indicators; complemented by binding workflows.

  • Canary deployments

    Technical
    Effect
    Detective, Impact-limiting
    Implementation level
    Model & training, Use & operations
    Classification and rationale

    Primarily technical: Software or analytical tools systematically produce and evaluate measurements, deviations, or attack indicators.

  • Feedback loops from production

    Organizational & process-based
    Effect
    Detective, Corrective
    Implementation level
    Model & training, Use & operations
    Classification and rationale

    Complementary control typeTechnical, People & competence

    Primarily organizational and process-based: A defined reporting, triage, and handling workflow turns observations into traceable follow-up actions; complemented by technical implementation as well as human expertise and judgment.

Framework mappings

Verified locations in OWASP, NIST AI RMF, MITRE ATLAS, the EU AI Act and further frameworks. The mappings are taxonomic, not evidence of compliance.

OWASP LLM Top 10 LLM07:2026NIST AI RMF Section 2.6 · MEASURE 2.5EU AI Act Article 26(5) · Article 55(1)(b)BSI R9BIML BIML-LLM raw:9 · BIML78 alg:1 · BIML78 eval:5

Verified references (9)

Every reference states the framework, the exact location and the publishing organisation.

EU AI Act obligations related to this threat

The obligations from the obligations compass whose articles the catalogue maps to this threat. Each leads to its entry with core, action step and evidence; whether it applies to you depends on role and risk class.

  • Monitor operation and suspend on risk Art. 26 Abs. 5 · applies from Annex III: 2 December 2027 · Annex I: 2 August 2028

    Statutory roles: Deployer. Classes: High-risk under Annex III, High-risk as a product or safety component.

    This reference is a review prompt. A threat match alone does not establish applicability.

  • Evaluate and mitigate systemic risks Art. 55 · applies from 2 August 2025

    Statutory roles: Provider of a general-purpose AI model. Classes: Model with systemic risk.

    This reference is a review prompt. A threat match alone does not establish applicability.

All 62 obligations by role, risk class and deadline in the obligations compass

More entries from the topic group Reliability and Responsible Use.

Overreliance People tend to accept convincing-sounding AI outputs without checking them (automation bias). Factual Inconsistencies (Hallucinations) AI language models sometimes produce content that sounds convincing but is simply made up (hallucinations), including citations, figures, … Shadow AI (Unsanctioned AI Service Use) Employees or business units use AI services without company approval or oversight (shadow AI).

Matching use cases

This list shows where to review this threat, rather than how high the risk is.

Belongs on the checklist for every AI system. What to review for every AI system

Particularly important for:

  • Credit scoring and creditworthiness assessmentThe model knows earlier years' payment behaviour. Changing interest rates, labour markets or customers make default estimates too high or too low, without immediate detection.
  • Payment fraud detectionNew scams, such as transfers following a call from a fake bank employee, differ from learned cases. Detection is poorer until enough new examples are available.
  • Money laundering detection and sanctions screeningNew laundering methods and payment channels are absent from training data. The model treats them as ordinary until retrained on new cases.
  • Life and health insurance risk assessment and pricingThe model knows earlier cohorts' mortality and illness patterns. Better treatment or a changed applicant population makes risk classes inaccurate.
  • Automated property valuation (AVM)The model learns from sales before the valuation date. Rapid price changes in particular locations make estimates inaccurate until new sales data is available.
  • Intelligent video analytics in public spacesLight, weather, seasons and alterations change camera images. The model recognises learned patterns less well without issuing a warning.
  • AI-assisted medical image analysisNew X-ray equipment, different acquisition settings or a different patient group changes images without informing the model. Detection performance silently falls below the level used for product assessment.
  • Clinical decision support (such as early sepsis detection)A model developed elsewhere encounters different patients, workflows and documentation practices and performs worse. At Michigan Medicine, measured performance fell well short of the manufacturer's claims.
  • AI support for emergency call handlingA model trained on one centre's calls performs worse with different languages, dialects or questioning practices. Pilots in France, Italy and Scotland therefore first trained it on local recordings.
  • Predictive maintenanceModifications, replaced parts or frost alter measurements without implying damage. Until adapted, the model issues false alarms or misses real deviations.
  • Visual and acoustic quality inspectionA new variant, supplier or lighting changes recordings without model retraining. The inspection then passes defects or rejects good parts.
  • Sales and demand forecastingWhen weather, promotions or a new competitor change demand, the learned patterns no longer fit. Forecasts can be wrong for weeks before anyone notices.
  • Automated driving and driver assistancePerception is familiar only with the regions, seasons and traffic conditions used in training. Snow, unfamiliar roadworks or unexpected behaviour by others impair obstacle recognition, and this is not immediately apparent.

Frequently asked questions

What is Model Drift & Degradation?
The quality of an AI model gradually declines in operation because the world keeps changing while the model stays frozen on old knowledge (model drift). Without dedicated monitoring, decisions get silently worse and nobody notices.
Which mitigations help against Model Drift & Degradation?
The catalogue lists 7 mitigations, each with control type, effect and implementation level. First among them are Continuous performance monitoring, Automated drift detection and Periodic retraining schedules.
Which EU AI Act obligations does Model Drift & Degradation touch?
The catalogue maps this threat to Art. 26 Abs. 5 and Art. 55 of the EU AI Act. In the obligations compass this leads to: Monitor operation and suspend on risk and Evaluate and mitigate systemic risks. Whether an obligation applies depends on role, risk class and deadline.

Assess this threat in your own system

The live demo contains all 52 threats of this catalogue, including the EU AI Act and GDPR assessment. The free single modules cover AI risk, the EU AI Act and GDPR. No sign-up; the assessment runs locally in your browser.

Start the live demo See the free modules

Cite this entry

For reports, policies or internal documents; the link leads directly to this entry.

“Model Drift & Degradation”. Versatile AI Risk Assessment, AI threat catalogue, as of October 2026.
https://www.versatile-ai-risk-assessment.com/en/wissensbasis/threats/model-drift/

← Back to the full catalogue