AI Regulation & Policy — 2026年8月3日 週次レポート
AI Regulation & Policyのニュース＆アップデート — すべての記述に一次ソースのリンク付き。
重要な発見
エグゼクティブサマリー（5件）
- •August 2, 2026 is the most consequential near-term date in global AI regulation: the EU AI Act's transparency obligations became enforceable, the Digital Omnibus entered into force, and 190 organisations signed the AI-generated content Code of Practice — collectively marking the EU's transition from the world's most ambitious AI regulatory framework on paper to one with active enforcement teeth.
- •The EU is simultaneously tightening its regulatory grip and opening its industrial wallet: the same week that AI Act enforcement launched, the Commission announced an AI Gigafactories call to unlock more than €30 billion in investment, confirming that Brussels views regulatory rigor and industrial AI leadership as complementary rather than competing objectives.
- •The U.S. state AI patchwork deepened further with Colorado's AI law revision and California's AI Transparency Act compliance guidance, while federal harmonization remained absent — organizations operating across multiple U.S. states face a structurally fragmented compliance landscape that will not simplify in the near term.
- •Japan's Cabinet approval of the Second Phase AI Basic Plan and the UK's Ofcom Phase 3 Online Safety Act implementation confirm that AI governance formalization is now a global phenomenon, with major economies outside the EU and U.S. establishing their own binding frameworks that create additional compliance obligations for internationally operating AI deployers.
- •The legal AI market's consolidation — Legora's fifth acquisition, Harvey's institutional investor backing, Thomson Reuters' LLM benchmarking — signals that the competitive landscape is narrowing around a small number of well-capitalized platform players, compressing the market for standalone AI governance overlays and raising the stakes for vendors whose differentiation rests on regulatory coverage breadth.
今回の要点（12件）
- 1.The EU Digital Omnibus on AI entered into force on July 27, 2026, and the Commission announced enforcement of AI Act transparency rules beginning August 2, 2026 — marking the end of the EU AI Act's preparatory phase and the start of its enforcement era. [6]
- 2.Approximately 190 organisations signed the Code of Practice on Transparency of AI-generated Content ahead of the legal obligations taking effect on August 2, 2026. [6]
- 3.The European Commission published EU AI Act Transparency Guidelines on July 20, 2026 to assist providers and deployers in complying with transparency requirements, and published practical guidance on the Cyber Resilience Act on July 27, 2026. [5]
- 4.The EU launched an AI Gigafactories call on July 30, 2026 to boost Europe's computing capacity and unlock more than €30 billion in investment, confirming a dual EU strategy of regulatory enforcement and industrial AI investment. [6]
- 5.Colorado revised its AI law and the California AI Transparency Act's compliance steps were published, continuing the U.S. state AI legislative expansion with no federal harmonization in sight. [1]
- 6.Japan's Cabinet approved the Second Phase AI Basic Plan on July 14, 2026, formalizing Japan's national AI governance architecture under its AI Act. [11]
- 7.Ofcom published its Phase 3 Online Safety Act implementation package on July 10, 2026, formally designating Category 1, 2A, and 2B services and publishing additional duties covering AI-generated content and deepfakes. [2]
- 8.The EDPB adopted Guidelines 03/2026 on web scraping in the context of generative AI for public consultation on July 7, 2026, extending EU regulatory reach into AI training data pipelines. [4]
- 9.Legora completed its fifth acquisition since March 2026 (Wexler), Harvey announced investments from JP Morgan and Goldman Sachs funds, and Thomson Reuters benchmarked a forthcoming LLM against Anthropic, OpenAI and Google models — confirming legal AI market consolidation. [8] [9]
- 10.OneTrust was named a Visionary in the 2026 Gartner Magic Quadrant for AI Governance Platforms, providing external market validation for the AI governance platform category. [1]
- 11.Tech Policy Press reported on July 30 that Brussels gained new AI Act enforcement powers as autonomous AI tests regulators, with the OpenAI-Hugging Face incident prompting calls for urgent Congressional oversight. [3]
- 12.An Illinois appeals court imposed higher fines to deter AI hallucinations in legal filings, signaling that courts are beginning to impose enterprise-grade AI accountability requirements. [8]
市場動向
EU AI Act Enters Active Enforcement Phase as Transparency Rules Take Effect
The week of July 27–August 2, 2026 marked a decisive inflection point for the EU AI Act: the Digital Omnibus on AI entered into force on July 27, the Commission announced it would begin enforcing AI Act rules and new transparency requirements on August 2, and approximately 190 organisations signed the Code of Practice on Transparency of AI-generated Content ahead of the legal obligations taking effect [6]. Tech Policy Press reported on July 30 that Brussels gained new AI Act enforcement powers o…
U.S. State AI Legislation Accelerates as Federal Vacuum Persists
The reporting period saw continued U.S. state-level AI legislative activity with no federal resolution in sight. Colorado revised its AI law, changing compliance expectations for businesses, and the California AI Transparency Act's requirements, deadlines, and compliance steps were published as new compliance guidance [1]. Law360 reported that federal and state lawmakers are expected to keep pressing efforts for companies to do more to keep children safe online and rein in potentially harmful us…
AI Governance Tooling Demand Institutionalizes Across Enterprise and Legal Sectors
Multiple signals this week confirmed AI governance is now a mainstream enterprise priority rather than a compliance edge case. OneTrust was named a Visionary in the 2026 Gartner Magic Quadrant for AI Governance Platforms (company announcement — may reflect promotional framing), with the blog noting that 'AI Governance is becoming a foundational business capability' [1]. Law.com reported that Thomson Reuters provided benchmarking data for a forthcoming LLM tested against models from Anthropic, Op…
競合動向
Legal AI Market Consolidation Continues Through Acquisition and Platform Deepening
Legora completed its fifth acquisition since March 2026 by purchasing Wexler, a fact intelligence platform for litigation, with Wexler's engineering group forming the founding team of Legora's London engineering hub [9] [8]. Harvey announced investments from JP Morgan and Goldman Sachs funds, with Goldman Sachs' contribution described as one of several investments the bank has made in legal tech since 2025 [8]. Thomson Reuters provided benchmarking data for a forthcoming LLM to be integrated int…
OneTrust Expands AI Governance Content Footprint Amid Regulatory Acceleration
OneTrust (company announcement — may reflect promotional framing) added multiple new AI governance articles this week, including coverage of Colorado's revised AI law, the California AI Transparency Act, and a CEO letter marking six months of operations, growing its article count from 657 to 660 across the reporting period [1]. The blog's sustained content expansion — translating each new regulatory development into compliance guidance — reinforces its competitive positioning as a regulatory cov…
AI Insurance and Liability Products Emerge as New Competitive Frontier
NewMod law firm Crosby announced it would provide professional liability insurance for its AI agents to enable autonomous legal work, described by Artificial Lawyer as a novel development in the legal AI market [9]. This development — the first reported instance of a law firm insuring its AI agents for legal liability — signals that the competitive frontier in legal AI is shifting from capability to accountability infrastructure, with liability coverage becoming a potential differentiator for fi…
制度・規制動向
EU AI Act Transparency Obligations Now Operative: Enforcement Begins August 2, 2026
The most significant regulatory development of the reporting period is the EU AI Act's transition from preparation to active enforcement. The Digital Omnibus on AI entered into force on July 27, 2026, and the Commission announced on July 31 that it would start enforcing AI Act rules and new transparency requirements on August 2, 2026 [6]. Approximately 190 organisations signed the Code of Practice on Transparency of AI-generated Content ahead of the legal obligations [6]. The Hunton Privacy Blog…
EU Cyber Resilience Act Guidance Published, Completing Near-Term EU Digital Compliance Stack
On July 27, 2026, the European Commission published practical guidance on the application of the Cyber Resilience Act, providing manufacturers, developers and businesses with guidance on mandatory cybersecurity requirements and reporting obligations [5] [6]. This guidance, arriving in the same week as the AI Omnibus entering into force, means organizations now face simultaneously operative compliance obligations under the EU AI Act transparency rules, the Cyber Resilience Act, and the ongoing NI…
Colorado AI Law Revision and California AI Transparency Act Add to U.S. State Compliance Burden
Colorado revised its AI law, changing compliance expectations for businesses, as reported by OneTrust on July 28, 2026 [1]. The California AI Transparency Act's requirements, deadlines, and compliance steps were published as new compliance guidance on July 30, 2026 [1]. Global Policy Watch reported that Illinois Governor JB Pritzker signed SB 315 on July 6, 2026, making Illinois the third state to enact comprehensive frontier AI safety requirements, with the law taking effect January 1, 2027 and…
EDPB Web Scraping Guidelines for Generative AI Open for Public Consultation
The European Data Protection Board adopted Guidelines 03/2026 on web scraping in the context of generative AI for public consultation on July 7, 2026, with analysis published by Privacy World Blog on July 29, 2026 [4]. The guidelines address the limits of GDPR orthodoxy in regulating AI training data collection, with implications for organizations using web-scraped data to train generative AI models. This development extends the EU's regulatory reach into the AI training data pipeline, complemen…
EU AI Gigafactories Call Launched, Signaling Industrial Policy Complement to Regulatory Framework
On July 30, 2026, the EU launched an AI Gigafactories call to boost Europe's computing capacity and unlock more than €30 billion in investment [6]. This industrial policy initiative — arriving in the same week as the AI Act's enforcement launch — confirms that the EU is pursuing a dual strategy of regulatory constraint and industrial investment simultaneously. The Gigafactories call is directly relevant to AI governance because it shapes the infrastructure on which compliant EU AI systems will b…
Japan Adopts Second-Phase AI Basic Plan, Formalizing National AI Governance Architecture
Japan's Cabinet Office announced on July 14, 2026 that the Second Phase AI Basic Plan (人工知能基本計画（第Ⅱ期）) was approved by the Cabinet, following the Fifth AI Strategy Headquarters meeting on July 10, 2026 [11]. This represents a significant formalization of Japan's national AI governance architecture under the AI Act (人工知能関連技術の研究開発及び活用の推進に関する法律), establishing a second-generation strategic framework that will shape AI research, development, and deployment policy across Japanese government and industr…
UK Online Safety Act Phase 3 Implementation: Ofcom Designates Category 1 Services and Publishes Additional Duties
On July 10, 2026, Ofcom published a package of draft materials as part of the third phase of its implementation of the Online Safety Act, including the long-awaited Register of Categorised Services formally designating services across Category 1, Category 2A, and Category 2B, draft Fraudulent Advertising Codes of Practice, and draft Additional Duties Code of Practice for Category 1 services covering user empowerment, identity verification, and freedom-of-expression assessments [2]. This phase 3 …
ソース活動
先週からの変化
EU AI Act Enforcement Begins August 2, 2026 — Transparency Rules Now Operative
The EU Digital Omnibus on AI entered into force on July 27, 2026, and the Commission announced on July 31 that enforcement of AI Act rules and new transparency requirements would begin on August 2, 2026. Approximately 190 organisations signed the Code of Practice on Transparency of AI-generated Content ahead of the legal obligations. [6] [5]
EU Cyber Resilience Act Practical Guidance Published
On July 27, 2026, the European Commission published practical guidance on the application of the Cyber Resilience Act, providing manufacturers, developers and businesses with guidance on mandatory cybersecurity requirements and reporting obligations — completing the near-term EU digital compliance stack alongside the AI Act transparency rules. [6] [5]
EU AI Gigafactories Call Launched to Unlock €30 Billion in Investment
On July 30, 2026, the EU launched an AI Gigafactories call to boost Europe's computing capacity and unlock more than €30 billion in investment, signaling that the EU is pursuing industrial AI investment policy in parallel with its regulatory enforcement framework. [6]
Colorado AI Law Revised and California AI Transparency Act Compliance Steps Published
Colorado revised its AI law, changing compliance expectations for businesses (reported July 28, 2026), and the California AI Transparency Act's requirements, deadlines, and compliance steps were published as new compliance guidance (reported July 30, 2026), adding to the U.S. state AI compliance burden. [1]
Japan Second-Phase AI Basic Plan Approved by Cabinet
Japan's Cabinet approved the Second Phase AI Basic Plan (人工知能基本計画（第Ⅱ期）) on July 14, 2026, following the Fifth AI Strategy Headquarters meeting on July 10, 2026, formalizing Japan's second-generation national AI governance architecture under the AI Act. [11]
ウォッチリスト — 今後の締切
EU AI Act Prohibition 9 (AI systems generating non-consensual sexually explicit content) takes effect, introduced as part of the AI Omnibus package
ソース: EU AI ActIllinois Artificial Intelligence Safety Measures Act (SB 315) takes effect for frontier AI developers
ソース: Hunton Privacy BlogEU AI Act high-risk AI system obligations take effect — strict requirements for risk assessment, data quality, logging, documentation, human oversight, and cybersecurity
ソース: EU AI ActIllinois SB 315 transparency-reporting and audit obligations begin for frontier AI developers
ソース: Hunton Privacy Blog示唆・見るべき論点（10件）
- 1.Organizations that have been treating EU AI Act transparency obligations as a future planning item must immediately shift to remediation mode: enforcement began August 2, 2026, and the Commission's simultaneous publication of transparency guidelines, the Cyber Resilience Act guidance, and the AI Gigafactories call signals that Brussels is operating all three levers — enforcement, guidance, and investment — simultaneously. [6]
- 2.The 190 organisations that signed the Code of Practice on Transparency of AI-generated Content before the August 2 deadline have established a compliance baseline that will be used as a reference point in enforcement proceedings against non-signatories — organizations that have not yet signed should assess their exposure immediately. [6]
- 3.The EU's AI Gigafactories call to unlock more than €30 billion in investment creates a strategic opportunity for AI governance platform vendors: organizations seeking EU public procurement or investment opportunities will need to demonstrate compliance with the AI Act's transparency rules as a prerequisite, making governance certification a commercial differentiator rather than merely a compliance cost. [6]
- 4.The EDPB's web scraping guidelines for generative AI — open for public consultation since July 7, 2026 — will directly affect organizations using web-scraped training data for EU-deployed AI models. The consultation window is the last opportunity to shape these guidelines before they become binding; organizations with significant web-scraping-based AI training pipelines should submit comments. [4]
- 5.Japan's Second Phase AI Basic Plan approval signals that organizations with AI operations in Japan face a new generation of national governance requirements that will be implemented through the AI Act's framework — organizations that have not yet mapped their Japan AI operations to the new plan's requirements should initiate that assessment now. [11]
- 6.The Illinois appeals court's decision to impose higher fines for AI hallucinations in legal filings — combined with the court's statement that nothing seems to be working — signals a judicial escalation trajectory that will affect all organizations using AI in legal proceedings, not just law firms. In-house legal teams deploying AI for contract review, litigation support, or regulatory filings face increasing judicial scrutiny. [8]
- 7.Legora's fifth acquisition since March 2026 and Harvey's institutional investor backing from JP Morgan and Goldman Sachs funds confirm that the legal AI consolidation trend documented in prior periods is accelerating rather than plateauing — organizations evaluating legal AI vendors should factor acquisition risk and platform stability into their procurement decisions. [8] [9]
- 8.The OpenAI-Hugging Face incident — prompting calls for urgent Congressional oversight and a Tech Policy Press podcast on its geopolitical AI governance implications — introduces a new category of AI governance risk: security incidents involving AI model infrastructure that have both technical and geopolitical dimensions. Organizations relying on third-party AI model providers should assess their supply chain security posture. [3]
- 9.Colorado's AI law revision and California's AI Transparency Act compliance guidance arriving in the same week confirm that U.S. state AI legislation is entering a revision cycle — not just an enactment cycle. Organizations that built compliance programs around the original versions of these laws must now assess whether their programs remain current, adding ongoing maintenance costs to the already significant initial compliance investment. [1]
- 10.The UK's Ofcom Phase 3 Online Safety Act implementation — formally designating Category 1 services and publishing additional duties covering AI-generated content — means that AI-powered platforms with UK users now face binding obligations on deepfake labelling, identity verification, and freedom-of-expression assessments that are distinct from and in addition to EU AI Act requirements. [2]
信頼度サマリー
今週引用したソース 13 件あなたが選んだ 30 件の監視URLから検出（1つのURLから複数記事が出ることがあります）。
各ソースは信頼度レベルに応じて重み付けされています。単独ソースの主張は AI 合成時に未検証としてフラグ付けされます。
参照ソース一覧
Source for OneTrust Gartner Visionary designation, Colorado AI law revision, California AI Transparency Act compliance steps, EU AI Act timeline analysis, and AI governance content expansion. Company announcement — may reflect promotional framing.
Source for Illinois SB 315 frontier AI safety law, UK Ofcom Phase 3 Online Safety Act implementation, EU tech sovereignty package analysis, and U.S. tech legislative and regulatory Q2 2026 update.
Source for EU AI Act enforcement powers activation, OpenAI-Hugging Face incident geopolitical implications, UK tech governance analysis, autonomous AI regulatory challenges, and open-source AI safety debate.
Source for EDPB Guidelines 03/2026 on web scraping in the context of generative AI, adopted for public consultation on July 7, 2026, and EU AI Act labelling obligations analysis.
Source for EU Digital Omnibus on AI entering into force on July 27, EU AI Act Transparency Guidelines published July 20, European Commission Cyber Resilience Act guidance published July 27, Illinois frontier AI model law, and CISA cyber incident reporting timeline.
Primary source for AI Omnibus entering into force July 27, Commission enforcement announcement for August 2, Code of Practice on Transparency of AI-generated Content with 190 signatories, AI Gigafactories call to unlock €30 billion, EU AI Act transparency guidelines, and Cyber Resilience Act guidance.
Source for EU AI Act risk-based framework details, transparency rules effective August 2026, AI Omnibus package provisions including Prohibition 9 effective December 2026, and high-risk AI system obligations effective December 2027.
Source for Legora's acquisition of Wexler, Harvey's JP Morgan and Goldman Sachs investments, Thomson Reuters LLM benchmarking, Illinois appeals court AI hallucination fines, Reddit DMCA claims against Perplexity and SerpApi, and legal AI market consolidation.
Source for Legora's acquisition of Wexler as fifth deal since March 2026, Crosby AI agent professional liability insurance, LegalOn 100 AI workflows launch, and Harvey-Microsoft relationship deepening.
Source for New York bill to track AI-related job layoffs, U.S. state AI and privacy legislative developments, and legal AI tools moving into specialized workflows.
Source for Japan's Cabinet approval of the Second Phase AI Basic Plan on July 14, 2026, and the Fifth AI Strategy Headquarters meeting on July 10, 2026.
Source for AI for Good Global Summit highlighting resilient AI through energy-efficient innovation on July 31, 2026, and UNESCO's ongoing AI governance and ethics framework activities.
Source for Trump administration AI strategy claims including $2.7 trillion in tech and AI investment attracted, America's AI Action Plan, and Genesis Mission announcement of more than $5 billion for AI for Science on July 22, 2026.
AI Regulation & Policyを毎週、自動で監視
このレポートは一次ソースのみから生成されています。テーマとソースを選べば、引用付きレポートが毎週届きます。7日間無料トライアル・$33/月から。
無料トライアルを始める