aicomply.
Lesson12 minChapter 3 of 9

Systemic Risk Classification

How GPAI models are classified as presenting systemic risk.

Systemic Risk Classification (Article 51)

Learning Objectives

By the end of this chapter, you will be able to:

  • Define "systemic risk" under the AI Act framework
  • Apply the computational threshold (10^25 FLOPS) to classify GPAI models
  • Identify factors that may trigger Commission designation
  • Execute notification obligations to the AI Office
  • Monitor models for approaching systemic risk thresholds
  • Understand the consequences of systemic risk classification

Article 51 introduces a critical distinction within the GPAI framework: models presenting "systemic risk" face significantly enhanced obligations. This tiered approach reflects the EU's recognition that the most capable AI models may pose risks at a societal scale.

Understanding Systemic Risk

Article 51(1) Classification Criteria

A GPAI model is classified as presenting systemic risk if it meets either of two conditions:

  1. Article 51(1)(a): It has high-impact capabilities evaluated on the basis of appropriate technical tools and methodologies, including indicators and benchmarks
  2. Article 51(1)(b): The Commission designates it as such, based on criteria equivalent to those set out in point (a), in accordance with Annex XIII, either ex officio or following a qualified alert from the scientific panel pursuant to Article 90(1)(a)

⚠️ Note: Article 51(1) itself does not define "systemic risk" by listing risk categories. The concept of high-impact capabilities and their potential effects is elaborated in Annex XIII and supporting Recitals, not in Article 51(1) directly.

Annex XIII criteria for assessing high-impact capabilities include:

CriterionAssessment Factors
Number of parametersModel size and complexity
Quality and size of datasetBreadth, depth, and curation of training data
Input/output modalitiesText, image, audio, video, code capabilities
Benchmarks and evaluationsPerformance on capability evaluations
Reach and scaleNumber of users, integrations, deployments
High-impact capabilitiesAssessed dangerous or transformative capabilities

"High-Impact Capabilities" Interpretation

IndicatorAssessment Criteria
Scale of deploymentMillions of users, broad integration
Capability breadthWide range of sophisticated tasks
Reasoning abilityComplex multi-step reasoning
Autonomous actionAbility to act with minimal human direction
Knowledge synthesisCombining information across domains
Code generationCreating functional software, exploits
PersuasionSophisticated content generation
MultimodalityIntegration of text, image, audio, video

The 10^25 FLOPS Threshold (Article 51(2))

Computational Threshold Rule

Article 51(2) establishes: A GPAI model shall be presumed to have high-impact capabilities when the cumulative amount of computation used for its training measured in floating point operations (FLOPs) is greater than 10^25.

Understanding the Threshold

Scale ReferenceFLOPSStatus
Small models (7B parameters)~10^22Well below threshold
Medium models (70B parameters)~10^23-10^24Below threshold
Large models (2024 frontier)~10^24-10^25Near threshold
Systemic risk threshold10^25Presumed systemic risk
Future frontier models>10^25Clearly above threshold

What Counts as "Training Compute"?

IncludedNot Included
Initial pre-trainingInference at deployment
Fine-tuning (if substantial)Minor adaptation/prompting
RLHF trainingDownstream provider fine-tuning
Multi-stage trainingEvaluation and testing
Distillation (from scratch)User interactions

Calculating Training Compute

Standard Estimation Formula:

Training FLOPS ≈ 6 × Number of Parameters × Number of Training Tokens

Example Calculation:

  • 70B parameter model trained on 2T tokens
  • 6 × 70×10^9 × 2×10^12 = 8.4×10^23 FLOPS
  • Below 10^25 threshold

💡 Expert Note: The 6× multiplier accounts for forward and backward passes plus optimizer operations. Actual compute may vary based on architecture and training approach.

Commission Designation (Article 51(1)(b))

Alternative Classification Path

Even if a model does not exceed 10^25 FLOPS, the Commission may designate it as presenting systemic risk based on:

CriterionAssessment Factors
Number of parametersModel size and complexity
Quality of datasetBreadth, depth, and curation of training data
Size of datasetVolume of training data
Input/output modalitiesText, image, audio, video, code capabilities
BenchmarksPerformance on capability evaluations
ReachNumber of users, integrations, deployments
Number of registered usersBusiness scale and market penetration
High-impact capabilitiesAssessed dangerous or transformative capabilities

Designation Process

StepActorAction
1Scientific PanelIssues qualified alert identifying potential systemic risk (Article 90(1)(a)), OR Commission acts ex officio
2CommissionInitiates investigation and evidence gathering based on Annex XIII criteria
3CommissionIssues designation decision
4ProviderMay challenge designation (Article 52(5))

Challenging a Designation

Article 52(2) provides that providers may, at the time of notification, present sufficiently substantiated arguments that their model does not present systemic risk and should not be classified as such. (Note: Article 52(5) addresses a separate process for reassessment of Commission designations.) Evidence may include:

  • Independent capability evaluations
  • Safety testing results
  • Limitation demonstrations
  • Use case restrictions
  • Technical safeguards implemented

Notification Requirements (Article 52)

Mandatory Notification

Article 52(1) requires GPAI providers to notify the Commission:

TriggerTimeline
Model meets 10^25 FLOPS thresholdWithin 2 weeks of meeting threshold
Reasonable grounds to believe threshold will be metBefore training completion
Commission designation receivedImmediate acknowledgment required

Notification Content

Article 52(1) requires that the notification include the "information necessary to demonstrate that the relevant requirement has been met." The following table distinguishes between the statutory minimum and recommended best-practice content:

ElementContent RequiredStatus
Information demonstrating the requirement is metEvidence that the 10^25 FLOPS threshold has been reached or will be reachedStatutory minimum (Article 52(1))
Provider identificationLegal entity, contact details, authorised representativeRecommended best practice
Model identificationName, version, release dateRecommended best practice
Training computeCumulative FLOPS calculation and methodologyRecommended best practice
Capability assessmentKnown capabilities and limitationsRecommended best practice
Intended distributionMarket placement plansRecommended best practice
Risk assessmentInitial systemic risk assessmentRecommended best practice

Notification Template

GPAI SYSTEMIC RISK NOTIFICATION
(Article 52, Regulation (EU) 2024/1689)

1. PROVIDER INFORMATION
   [Legal name, address, contact, authorised representative]

2. MODEL IDENTIFICATION
   [Model name, version, planned release date]

3. TRAINING COMPUTE
   Total FLOPS: [X.XX × 10^YY]
   Calculation methodology: [Description]
   Training completion date: [Date]

4. CAPABILITY ASSESSMENT
   [Summary of model capabilities]
   [Known limitations]

5. RISK ASSESSMENT
   [Identified systemic risks]
   [Planned mitigations]

6. DISTRIBUTION PLANS
   [Intended market placement approach]
   [Timeline]

Submitted by: [Name, title]
Date: [Date]

Consequences of Systemic Risk Classification

Enhanced Obligations

Baseline GPAI+ Systemic Risk Additions
Technical documentation+ Model evaluation and adversarial testing
Downstream information+ Systemic risk assessment at Union level
Copyright policy+ Incident tracking and reporting
Training data summary+ Cybersecurity protection

Regulatory Scrutiny

AspectStandard GPAISystemic Risk GPAI
AI Office oversightGeneralEnhanced monitoring
Evaluation requestsAd hocRegular requirements possible
Incident reportingVia downstream providersDirect to AI Office
Enforcement focusDocumentationActive risk management

Monitoring Approaching Threshold

Internal Monitoring Framework

MetricMonitoring FrequencyThreshold Alert
Training FLOPS accumulatedDaily during training10^24 (approaching)
Parameter countPer training iterationPlanning stage
Dataset size growthWeeklyMajor expansions
Capability evaluationsPer checkpointSignificant capability gains
User/deployment scaleMonthlyRapid growth

Pre-Threshold Preparation

For models approaching 10^25 FLOPS:

  • Begin Article 55 compliance preparation
  • Establish adversarial testing programme
  • Develop systemic risk assessment methodology
  • Implement enhanced incident tracking
  • Review cybersecurity measures
  • Prepare AI Office notification
  • Engage Scientific Panel proactively

What You Learned

Key concepts from this chapter

Systemic risk classification applies to GPAI models with "high-impact capabilities" affecting the EU market

10^25 FLOPS creates a **presumption** of systemic risk—below-threshold models may still be designated

Commission can designate models based on multiple factors beyond compute

Providers must notify the Commission within 2 weeks of meeting the threshold

Systemic risk classification triggers Article 55 enhanced obligations

Chapter Complete

GPAI Compliance

3/9

chapters