Skip to main content

AI Risk Management

Purpose

This document describes how the AI Conveyor identifies, assesses, and controls the risks of AI-based solutions: machine learning models, language models, knowledge assistants, automations, and agentic scenarios.

The aim of risk management is not to ban AI, but to prevent a situation where a solution affects a business process, customers, money, or regulatory obligations without clear quality, an owner, and a way to stop it.

Core ideas

  • Risk depends on the impact of the solution. An internal assistant and a solution that affects a customer credit limit require different levels of rigor.
  • Human in the loop is a management decision. You need to explicitly define where AI advises, where it prepares a draft, and where an action is performed automatically.
  • Quality must be measurable. You cannot launch a solution if it is unclear how the accuracy, completeness, stability, or usefulness of the result is verified.
  • Constraints must be visible to the user. If a solution is not suitable for legally significant conclusions, this must be explicitly recorded.
  • The stop must be described in advance. A working solution must have conditions for review, shutdown, or rollback.

How it works

AI risk management is embedded in the stage gates:

StageWhat is checkedDecision
Newwhether there is an obviously high risksend to assessment, reject, or request security
Assessmentrisk class, data, process impact, human in the loopchoose the delivery route and the level of control
Deliveryquality, testing, constraints, logging, rollbackallow launch or return for rework
Awaiting impactactual quality and impact on the processconfirm the impact, adjust, or stop
Supportdegradation, incidents, changes in data and processcontinue, reconsider, or shut down

Related sections: AI governance model, data governance, stage gate model.


Main types of risk

RiskWhat can go wrongHow to control it
Result errorthe solution gives a wrong answer, forecast, or recommendationtest sets, manual review, quality thresholds
Hallucinationsa language model confidently invents factsreferences to sources, context restriction, verification of critical answers
Biasthe result is systematically worse for some customers, products, or processessample analysis, fairness checks, usage constraints
Data leakagesensitive information ends up in an unsuitable environmentdata classification, masking, access control
Uncontrolled automatic actionthe solution performs an action that a human should confirmhuman in the loop, limits, action log
Quality degradationthe data or process changed, but the solution keeps working the old waymonitoring, periodic review, stop signals
Non-reproducibilityit is impossible to understand why a result was producedversions of data, model, prompt, rules, and parameters
Reputational riskthe solution looks unfair, opaque, or unsafeexplainability, usage rules, communication of constraints

Impact classes

Minimum classification by impact:

ClassDescriptionRequirements
HintAI helps the user but does not make the decisiona warning about constraints, basic quality check
DraftAI prepares a text, document, calculation, or recommendation for human reviewmandatory human confirmation, logging
RecommendationAI influences a choice, priority, or actionquality metrics, explanation of factors, decision owner
Automatic actionAI triggers an action without manual confirmationstrict limits, rollback, audit, extended sign-off
Critical impactthe solution affects money, customers, risk, security, or legally significant processesindependent review, human in the loop, regular review

The higher the impact class, the more requirements there are for verification, monitoring, and documentation.


What to record on the initiative card

For initiatives with medium and high risk, you need to record:

  • the impact class;
  • the type of decision;
  • the data used;
  • the business decision owner;
  • the technical solution owner;
  • usage constraints;
  • quality metrics;
  • verification results;
  • the mode of human involvement;
  • stop conditions;
  • the rollback plan;
  • the date of the next review.

This data can be stored as card fields, artifacts, or tasks, but it must be available when passing through the stage gates.


Pre-launch check

Before moving an initiative to awaiting impact, you need to answer the questions:

  • on what data the solution was tested;
  • which quality metrics were achieved;
  • what counts as an error;
  • who reviewed the results;
  • which constraints are known;
  • what the user will see in the interface;
  • who is responsible for incidents;
  • how to shut down the solution;
  • what happens if the model, data, or external service is unavailable.

If there are no answers to these questions, the solution may be a prototype, but not a governed adoption.


Post-launch monitoring

After launch you need to track:

  • result quality;
  • usage volume;
  • the share of manual corrections;
  • complaints and incidents;
  • changes in input data;
  • integration latency and errors;
  • the cost of running the solution;
  • the actual impact on business metrics.

For high-impact solutions, a review must be set: for example, once a quarter, after a data source changes, or after a significant incident.


The role of the AI assistant

The AI assistant can:

  • assemble a description of the risks;
  • prepare a draft security review;
  • point out constraints;
  • create a list of questions for the data owner;
  • help with the initiative report;
  • create quality check tasks.

But the AI assistant must not approve risky transitions on its own, give the final security sign-off, or bypass access rights.


Stop conditions

A solution must be stopped or reviewed if:

  • quality is below the agreed threshold;
  • a leak or access violation is detected;
  • the data source has changed;
  • users are correcting the result en masse;
  • the operating cost has grown;
  • the solution is used for unintended purposes;
  • the business impact is not confirmed;
  • new regulatory or internal constraints have appeared.

Anti-patterns

Bad risk management:

  • "we'll check quality after launch";
  • a language model answers without references to sources in a critical process;
  • automation performs an action without a log and rollback;
  • constraints are written in a document but not visible to the user;
  • there is no incident owner;
  • there is no review date;
  • risk does not affect the transition decision.

Good risk management makes risk part of the initiative's route, not a late emergency brake.