EU AI法における汎用AIに関する義務：第5章ガイド
EU AI法の第5章は、汎用AI（GPAI）を規制します。これは、多くの人が基盤モデルと呼ぶものです。2つのレベルがあります。第53条に基づく標準的なGPAI義務と、第55条に基づくシステム的リスクを持つモデルに対する追加の義務です。
EU AI法における汎用AIに関する義務：第5章ガイド
EU AI法の第5章は、汎用AI（GPAI）を規制します。これは、多くの人が基盤モデルと呼ぶものです。2つのレベルがあります。第53条に基づく標準的なGPAI義務と、第55条に基づくシステム的リスクを持つモデルに対する追加の義務です。
最終更新日: 2026年7月4日
GPAIとは何か？
EU AI法は、汎用AIモデルを次のように定義しています。
重要な汎用性を示し、市場にどのように投入されたかに関わらず、幅広い異なるタスクを適切に実行できるAIモデル。ほとんどの大規模言語モデル、画像生成モデル、およびマルチモーダル基盤モデルが該当します。
第53条：標準的なGPAIプロバイダーの義務
すべてのGPAIプロバイダーは、これらの基本的な要件を満たす必要があります。
- トレーニングおよびテストプロセスを含む技術ドキュメントを維持する
- モデルを統合する下流のプロバイダーに情報を提供する
- テキストおよびデータマイニングのオプトアウトを尊重するポリシーを含む、EUの著作権法を遵守する
- 十分に詳細なトレーニングデータの概要を公開する
- EU内に拠点を置いていない場合は、EU代表を指名する
第55条：システム的リスクGPAI義務
システム的リスクの閾値（現在は10²⁵ FLOPsのトレーニング計算量、参考値）を満たすモデルは、追加の義務を負います。
- 敵対的テストを含むモデル評価
- システム的リスクの評価と軽減
- サイバーセキュリティ保護
- 重大なインシデントをAIオフィスに報告する
- エネルギー消費量の開示
行動規範
委員会は、GPAIプロバイダー向けの自主的な行動規範を策定しました。
採用は任意ですが、それが対象とする義務に対する適合の推定を可能にします。
オープンソースの例外
オープンソースのGPAIモデルは、規制において有利な扱いを受けます。
自由およびオープンソースライセンスの下でリリースされたGPAIモデルは、著作権およびトレーニングデータ概要の義務を除き、第53条から大部分が免除されます。また、この例外は、システム的リスクレベルのモデルには一切適用されません。
GPAI Compliance Timeline
The Chapter V obligations phase in on their own schedule, distinct from the high-risk deadlines:
- 2 August 2025 — the GPAI provider obligations under Articles 53 and 55 begin to apply to models placed on the market from this date
- 2 August 2026 — the AI Office's supervisory and enforcement powers over GPAI providers become applicable
- 2 August 2027 — GPAI models already on the market before 2 August 2025 must be brought into compliance by this date
- Signing the Code of Practice — providers who adhere to the GPAI Code of Practice can rely on it to demonstrate conformity while harmonised standards are still being developed
Dates reflect Regulation (EU) 2024/1689 as it stands; confirm against the latest Official Journal text and AI Office guidance, which may evolve.
How AIAgentree helps
AIAgentree governs how your agents use GPAI models in production, giving you the downstream documentation and incident evidence Chapter V expects.
- Tamper-evident decision records log how a GPAI-backed agent reasoned toward each decision, giving you durable evidence for downstream-provider documentation and serious-incident reporting
- Human-oversight and approval workflows plus outcome tracking help you evidence the risk-mitigation and monitoring practices Article 55 asks of systemic-risk deployments
- Python and TypeScript SDKs with REST, MCP, A2A, and OpenTelemetry integrate model-usage evidence into your compliance stack, with EU data residency in Germany and audit-fit retention of at least six months
Frequently Asked Questions
What counts as a general-purpose AI model under the EU AI Act?
A GPAI model is one that displays significant generality and can competently perform a wide range of distinct tasks, regardless of how it is released. Most large language models, image-generation models, and multimodal foundation models fall within this definition.
What is the systemic-risk threshold for GPAI?
A GPAI model is presumed to carry systemic risk when the cumulative compute used for its training exceeds 10^25 floating-point operations, an indicative threshold the Commission can adjust. Such models face extra duties under Article 55, including model evaluations, adversarial testing, risk mitigation, cybersecurity, and serious-incident reporting.
Is the GPAI Code of Practice mandatory?
No. The Code of Practice is voluntary. Providers who adhere to it can use it to demonstrate conformity with their Chapter V obligations while harmonised standards are still being finalised; providers who do not sign it must demonstrate compliance by other adequate means.
Are open-source models exempt from the GPAI rules?
Only partly. GPAI models released under a free and open-source license are largely exempt from the Article 53 obligations, but they must still meet the copyright-policy and training-data-summary duties. The exemption does not apply at all to models that meet the systemic-risk threshold.
Continue exploring the EU AI Act guide
EU AI Act Compliance Guide
The complete guide to EU AI Act compliance for AI agents — start here.
Article 12 — Record-Keeping & Logging
What every high-risk AI system must log, and how to capture it.
Article 14 — Human Oversight
Designing effective human-in-the-loop controls for AI decisions.
Annex III — High-Risk AI Systems
Which AI use cases the Act classifies as high-risk.
EU AI Act Compliance Checklist
A step-by-step checklist to reach and document compliance.
Compliance Cost Calculator
Estimate your EU AI Act compliance effort and cost.
Deadlines & Timeline
Key enforcement dates, including the August 2, 2026 deadline.
Fines & Penalties
Penalty tiers up to €35M or 7% of global annual turnover.
Transparency Obligations (Art. 13 & 50)
Disclosure duties for AI systems and their outputs.
Risk Management & Conformity Assessment
Build a risk management system and assess conformity.
EU AI Act for US Companies
Extraterritorial scope and what US providers must do.
Omnibus Update
The latest changes to the EU AI Act timeline and rules.
Penalty Calculator
Estimate your maximum fine under the Article 99 tiers.
Article 11 + Annex IV
What technical documentation the EU AI Act requires.
Article 26: Deployer Obligations
What deployers of high-risk AI must do, including log retention.
Article 17: Quality Management
The QMS providers of high-risk AI must document.
Article 10: Data Governance
Data quality, bias mitigation, and governance duties.
Article 4: AI Literacy
The staff AI-literacy duty in force since February 2025.
Deployer vs Provider
Who bears which obligation — and when a deployer becomes a provider.
FRIA (Article 27)
Who must run a Fundamental Rights Impact Assessment, and how.
Who Does It Apply To?
Scope, operators, and the extraterritorial reach of the EU AI Act.
Post-Market Monitoring
Articles 72–73: ongoing monitoring and incident reporting.
ISO 42001 vs EU AI Act
How the voluntary standard and the binding law fit together.
NIST AI RMF vs EU AI Act
A practical crosswalk between the framework and the law.
EU AI Act for Healthcare
High-risk medical AI, MDR/IVDR interplay, and clinician oversight.
EU AI Act for Financial Services
Credit scoring, insurance pricing, and existing financial regulation.
EU AI Act for HR & Employment
Hiring AI as high-risk, plus NYC LL144 and EEOC overlap.