Providers of systemic-risk GPAI models must secure the model and its infrastructure

Practical action

Apply a documented security standard to weight storage, training clusters and inference endpoints, with penetration tests and insider-risk controls.

Who does it apply to?

Providers of general-purpose AI models classified as having systemic risk under Article 51.

Actors

General-purpose AI model provider

Sectors

Cross-sector / all sectors

Use cases

Generative AI and foundation models

Applies from: 2 August 2025

Which controls meet this duty?

Satisfies: the control, operated properly, does the work the duty asks for. Supports: it contributes but the duty needs more. Each control page lists every other duty it serves, so work done once can be counted once.

• satisfiesTechnical measureQuality or testing lead · at launch and on material change
  Accuracy, robustness, fairness and security testing

  Serves 15 recorded duties · evidence: Pre-release test report, Test plan and acceptance criteria, Release test sign-off

  Security testing of the model and its serving infrastructure.

• supportsPolicyHead of AI safety · annual
  Frontier model safety and security framework

  Serves 9 recorded duties · evidence: Published frontier safety framework, Dangerous-capability evaluation report, Threshold notification to an authority

  Framework commits to the weight-security measures.

Templates that cover this duty

Generated from the records. Free · no account · links emailed to your work address. This duty is cited in each file with its source reference and a link back here.

• EU AI Act Role and Risk Classifier

  Checklist · XLSX · v4

• EU AI Act Article 50 Transparency Kit

  Kit · DOCX and XLSX · v44

• NIST AI RMF ↔ EU AI Act ↔ ISO/IEC 42001 Crosswalk

  Crosswalk · XLSX · v5

• AI System Technical Documentation

  Kit · DOCX and XLSX · v3

All 7 templates that cover it →

What evidence would a reviewer expect?

Framework mappings

Original editorial crosswalks. They cite clause numbers only and reproduce no standard text; confidence reflects how direct the mapping is.

See every European Union duty mapped this way →

Frequently asked questions

Cite this record

AIPolicyTracker (2026). “Providers of systemic-risk GPAI models must secure the model and its infrastructure (EU AI Act)” [AIPT-OBL-0046, verified 26 September 2026]. https://aipolicytracker.org/id/AIPT-OBL-0046 (accessed 8 November 2026). Data licensed CC BY 4.0.

Permanent id AIPT-OBL-0046: it never changes or gets reused, and its address keeps working if this page moves.

Cite the official text alongside it: Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence, Official Journal of the European Union, https://eur-lex.europa.eu/eli/reg/2024/1689/oj.

@misc{aipolicytracker2026_providersofsystemicriskg,
  author       = {{AIPolicyTracker}},
  title        = {{Providers of systemic-risk GPAI models must secure the model and its infrastructure (EU AI Act)}},
  year         = {2026},
  howpublished = {\url{https://aipolicytracker.org/obligations/eu-ai-act-art-55-systemic-risk-cybersecurity}},
  url          = {https://aipolicytracker.org/obligations/eu-ai-act-art-55-systemic-risk-cybersecurity},
  urldate      = {2026-11-08},
  note         = {Data licensed CC BY 4.0. Accessed 8 November 2026}
}

Change history

Similar obligations in other instruments

• Achieve appropriate accuracy, robustness and cybersecurity — EU AI Act, European Union
• Employers and employment agencies must obtain an independent bias audit before using an automated employment decision tool — NYC Local Law 144 (automated employment decision tools), New York (United States)