Model Drift

Full definition

Drift comes in two flavors: data drift (input distribution changes — new product categories, post-COVID consumer behavior, new fraud patterns) and concept drift (the relationship between inputs and outputs changes — what counted as "spam" in 2018 vs 2025). Detection requires continuous monitoring of input statistics, output distributions, and ground-truth feedback. Mitigation is usually retraining, sometimes architectural changes.

Why it matters

Models that worked great in pilot fail silently in production. A small accuracy drift can erase much of a model's ROI in fraud detection, demand forecasting, or pricing. Drift monitoring is non-negotiable governance for any model in regulated production use; the EU AI Act requires providers of high-risk systems to run post-market monitoring that tracks performance throughout the system's lifetime (Article 72), once the high-risk rules apply: 2 December 2027 for Annex III systems and 2 August 2028 for Annex I systems under the Digital Omnibus (Regulation (EU) 2026/1744).

Example

A demand-forecasting model deployed in 2019 collapsed during COVID-19 because consumer behaviour shifted overnight outside its training distribution. Companies with active drift monitoring caught it in days; others discovered it only after stockouts and excess inventory hit P&L.

Next steps

• Get a free workflow map →We map one of your workflows and give you a written verdict: software, a process change or AI. No obligation.
• AI consulting cost guide →Hourly rates and project budgets by firm tier, from independents to the Big Four.

Related terms

• MLOps
  MLOps is the discipline of operating machine-learning and AI systems reliably in production — covering data pipelines, model training, deployment, monitoring, drift detection, governance, and incident response — analogous to DevOps for traditional software.
• Training Data
  Training data is the curated dataset used to teach a machine-learning model — every pattern the model can recognize, every bias it inherits, and every limit to its accuracy ultimately traces back to this data.
• AI Governance
  AI governance is the framework of policies, roles, controls, and processes an organization uses to ensure its AI systems are lawful, safe, fair, accountable, and aligned with business intent — across the full lifecycle from problem framing to retirement.
• AI Audit
  An AI audit is a structured, evidence-based examination of an AI system or AI program against defined criteria — covering training data, model, deployment context, monitoring, and governance — performed by an internal team, an external firm, or a regulator.
• Model
  An AI model is the trained artifact — a specific set of numerical weights plus an architecture — produced when a learning algorithm runs against training data, and the thing that actually gets deployed and audited.

More deployment & operations terms

• AI Pipeline
  An AI pipeline is the end-to-end sequence of stages that turns raw data into deployed AI predictions or content — ingestion, cleaning, feature engineering, training, evaluation, deployment, monitoring, and retraining — usually orchestrated as code.
• AI Agent / Agentic AI
  An AI agent is an LLM-driven system that, given a goal, plans a sequence of steps, calls tools or APIs, observes the results, and iterates until the goal is reached — going beyond single-turn chat to multi-step autonomous action.