EU AI Act Compliance Tracker 2026
Last updated: August 24, 2026 · Updated weekly · Next scheduled update: August 30, 2026 | Maintained by: Axis Intelligence Research & Sarah Mitchell
Quick Answer
The EU AI Act has been in enforcement for three weeks. Article 50 transparency obligations and the Commission’s GPAI penalty powers activated August 2, 2026. Regulation (EU) 2026/1744 is in force since July 27, 2026, shifting Annex III high-risk obligations to December 2, 2027. As of August 24, 2026 — 22 days into the enforcement era — no fine, formal investigation, or enforcement decision under the EU AI Act has been publicly confirmed by the European AI Office or any national competent authority. The AI Office has stated that technical compliance dialogues are its preferred initial tool. The Commission has confirmed that GPAI violations dating back to August 2025 are within enforcement scope.
What This Tracker Covers
Axis Intelligence Research monitors every material EU AI Act compliance milestone — enforcement dates, regulatory guidance releases, institutional developments, Digital Omnibus amendments, GPAI Code of Practice updates, and national competent authority actions. Each entry carries a primary source fetched and verified at entry time. An unsourced row does not ship.
This tracker is the fact table for Axis Intelligence Research’s monthly EU AI Act Compliance Bilans. Numbers in those bilans recompute from this dataset. Do not reconcile figures from the bilan against any other source — this tracker is the primary record.
What counts for inclusion: Official publications from the European Commission, European AI Office, European Parliament, Council of the EU, and national competent authorities. Formal enforcement actions with a public decision number or official press release. Binding or quasi-binding guidance (guidelines, codes of practice, implementing acts) from primary EU institutions. Announcements that materially change the compliance timeline or obligations for organizations operating in or serving the EU.
What does not count: Unverified media reports of enforcement actions without a primary source. Consultations that have not closed and produced formal output. Commentary, analysis, and opinion from law firms or compliance vendors.
How We Curate This List
Inclusion criteria: An event enters this tracker when it meets all three of the following: (1) it originates from a primary EU institution or national competent authority named above; (2) it materially affects compliance obligations, timelines, or enforcement risk for operators in scope; (3) it can be linked to a fetched primary-source document with a verified URL.
Definitions: Enforcement action means a formal investigation opened, a sanction imposed, or a market withdrawal ordered by a competent authority, with a public decision number or official press release. Milestone means a change to the Act’s application schedule, a new binding or quasi-binding guideline, or a code of practice publication that creates a compliance safe harbor. Regulatory action means formal guidance, consultation output, or implementing legislation that clarifies obligations.
Source discipline: Every entry links to the primary EU institution document — not a law firm summary, not a news article. Where the primary document requires login, the entry notes this and links to the closest public summary from the issuing institution.
Currency: The tracker is updated weekly. Entries are added within 48 hours of a primary-source document becoming public. Entries do not appear before a primary source exists.
Enforcement & Milestones — Reverse Chronological
August 24, 2026 — Enforcement status at week three: no formal action published; systemic risk scope confirmed
Type: Enforcement status — Axis Intelligence Research verification finding Status: No formal enforcement decision published as of this update
Twenty-two days into the EU AI Act enforcement era, no fine, formal investigation opened by Commission decision, or market withdrawal order has been publicly confirmed by the European AI Office, the European Data Protection Supervisor, or any national competent authority. Axis Intelligence Research verified this across the Commission presscorner, the AI Office page, the enforcement framework page, and the AI Act Service Desk resources on August 24, 2026. The absence is consistent with the AI Office’s stated enforcement posture: technical compliance dialogues precede formal escalation, and the Office has said it will use formal powers where dialogues prove insufficient. Two scope points from the Commission’s July 31 press release (IP/26/1714) merit recording. First, the Commission confirmed that GPAI obligation violations dating back to August 2, 2025 — when the substantive duties took effect — fall within enforcement scope, meaning the twelve-month gap between obligation and enforcement does not shield prior non-compliance. Second, the press release specifies the systemic risk categories that trigger additional obligations for the most advanced GPAI models: chemical, biological, radiological and nuclear incidents; loss of control; cyber offence; harmful manipulation; and threats to fundamental rights. The Commission explicitly noted that these categories address risks including European cybersecurity and AI acting outside human control.
Compliance deadline affected: None — status verification entry Primary source: European Commission press release IP/26/1714, July 31, 2026; Commission enforcement framework page; AI Office page (all verified August 24, 2026) Source URL: https://ec.europa.eu/commission/presscorner/detail/en/ip_26_1714 Retrieved: August 24, 2026
August 2, 2026 — EU AI Act enforcement active: GPAI penalties and Article 50 transparency fully operative
Type: Enforcement milestone — enforcement era begins Status: Active; no formal decision published as of August 24, 2026
The EU AI Act entered its enforcement era on August 2, 2026. The European Commission’s AI Office and national competent authorities activated full enforcement powers across three tracks simultaneously. First, the AI Office can now issue formal information requests (RFIs), conduct model evaluations with independent experts, require corrective measures, and impose fines of up to €15 million or 3% of global annual turnover on GPAI model providers under Article 101. Second, Article 50 transparency obligations are enforceable: chatbots and interactive AI systems must disclose AI identity at the start of interactions; deepfakes must be labelled; AI-generated or altered content must carry machine-readable marks. Legacy systems placed on the EU market before August 2 have until December 2, 2026, for the machine-readable marking obligation under the Digital Omnibus grace period. Third, the national penalty framework under Article 99 is fully operative — up to €35 million or 7% of global turnover for Article 5 prohibited practice violations. The AI Office has stated publicly that “technical compliance dialogues” will be its preferred initial approach before escalating to formal enforcement. Euractiv has reported bilateral engagement between the AI Office and GPAI providers; no formal investigation with a published case number or Commission decision has been confirmed as of this update. The AI Office is recruiting 40 additional contractual agents to strengthen enforcement capacity.
Compliance deadline affected: Enforcement active across all in-force obligations Primary source: European Commission enforcement framework page; Commission press release IP/26/1714, July 31, 2026; Wilson Sonsini analysis citing Commission enforcement powers Source URL: https://digital-strategy.ec.europa.eu/en/policies/enforcement-ai-act Retrieved: August 12, 2026
July 30, 2026 — Commission launches AI Gigafactories call: up to 7 facilities, €30 billion+ investment
Type: Policy milestone — AI infrastructure and evaluation capacity Status: Call for tenders launched; relevant to GPAI model evaluation capacity
The European Commission launched a call for tenders on July 30, 2026 to establish up to seven AI Gigafactories across Europe, part of its AI Continent Action Plan. The call is administered through the EuroHPC Joint Undertaking and unlocks more than €30 billion in investment. The Gigafactories initiative is directly relevant to EU AI Act enforcement: the Commission’s July 7, 2026 Action Plan on Cybersecurity and AI committed to launching a call to increase EU evaluation capacity for AI models before they are placed on the EU market, with full evaluation capacity expected by 2027. The Gigafactories will provide the compute infrastructure for the AI Office’s model evaluation powers under Article 92 of the AI Act. An online information session was held on June 16, 2026; call documents were published July 1.
Compliance deadline affected: Shapes GPAI model evaluation enforcement capacity from 2027 onward Primary source: European Commission press release, “EU launches AI Gigafactories call to boost Europe’s computing capacity and unlock more than €30 billion in investment,” July 30, 2026 Source URL: https://digital-strategy.ec.europa.eu/en/news/eu-launches-ai-gigafactories-call-boost-europes-computing-capacity-and-unlock-more-eu30-billion Retrieved: August 12, 2026
July 31, 2026 — Commission announces enforcement commencement on August 2; publishes complaint tools and >180 Article 50 CoP signatories
Type: Enforcement milestone — official enforcement start announcement Status: Enforcement active from August 2, 2026
The European Commission published an official press release on July 31, 2026, announcing that from August 2, 2026 the AI Office and national authorities will begin enforcing the AI Act. The release details that Article 50 transparency obligations activate simultaneously: chatbots and interactive AI systems must disclose that users are interacting with AI; deepfakes must be labelled; AI-generated or altered content must carry machine-readable marks. The Commission simultaneously published a first list of more than 180 organisations that have signed the Code of Practice on Transparency of AI-generated Content — confirming the Code’s operational status as a compliance safe harbor for Article 50. The Commission also published links to the AI Act complaints tool, AI Act Whistleblower Tool, and complaints channel for downstream providers using GPAI models, operationalising enforcement access for citizens and industry.
Compliance deadline affected: August 2, 2026 — Article 50 transparency obligations; August 2, 2026 — GPAI enforcement powers (Article 101) Primary source: European Commission press release, “Commission starts enforcing AI Act rules and new transparency requirements on 2 August,” July 31, 2026 Source URL: https://digital-strategy.ec.europa.eu/en/news/commission-starts-enforcing-ai-act-rules-and-new-transparency-requirements-2-august Retrieved: August 1, 2026
July 27, 2026 — Regulation (EU) 2026/1744 enters into force; Digital Omnibus amendments binding law
Type: Legislative milestone — binding entry into force Status: In force
Regulation (EU) 2026/1744 entered into force on July 27, 2026, three days after publication in the Official Journal of the European Union on July 24, 2026. The regulation is formally designated as the “Digital Omnibus on AI” (also referred to as Omnibus VII). Its entry into force, described in the regulation’s preamble as occurring “as a matter of urgency” given the proximity of the August 2, 2026 deadline, amends Regulation (EU) 2024/1689 (the EU AI Act), Regulation (EU) 2018/1139 (basic aviation), and Regulation (EU) 2023/1230 (machinery regulation). The amended EU AI Act timeline is now legally binding: Annex III high-risk stand-alone AI systems must comply by December 2, 2027; Annex I embedded high-risk AI by August 2, 2028; the nudification prohibition and watermarking obligations apply from December 2, 2026. The Article 5 prohibitions, GPAI obligations, and Article 50 transparency obligations remain on the original August 2, 2026 schedule and are unaffected by the Omnibus. This is the first formal amendment to the EU AI Act since its adoption in June 2024.
Compliance deadline affected: All Omnibus amendments now legally binding: Annex III high-risk December 2, 2027 · Annex I embedded August 2, 2028 · Nudification prohibition December 2, 2026 · Watermarking December 2, 2026 Primary source: Official Journal of the European Union, Regulation (EU) 2026/1744 of 8 July 2026; FASI analysis confirming OJ publication and entry into force dates Source URL: https://fasi.eu/en/articles/strategies/29051-digital-omnibus-package.html Retrieved: August 1, 2026
July 20, 2026 — Commission publishes final Article 50 guidelines on transparency obligations
Type: Regulatory action — quasi-binding interpretive guidance Status: Final guidelines published; applies from August 2, 2026
The European Commission published its final Guidelines on the transparency obligations under Article 50 of the EU AI Act on July 20, 2026, replacing the draft version published in May 2026. The 51-page guidelines are non-binding but will be followed by national market surveillance authorities and the AI Office in enforcement. Key clarifications in the final text: Article 50 applies to chatbots, generative AI tools, emotion recognition systems, biometric categorisation tools, and deepfake technologies regardless of high-risk classification; no retroactive labelling obligation — for image, audio and video deepfakes the relevant date is the date of generation (pre-August 2 content does not require retroactive marking); for text published on matters of public interest, the relevant date is the publication date (text generated before August 2 but published on or after must be labelled); a grace period under the Digital Omnibus applies to pre-existing generative AI systems placed on the market before August 2, 2026, which have until December 2, 2026 for the machine-readable marking obligation. Fines for non-compliance reach €15 million or 3% of global annual turnover.
Compliance deadline affected: August 2, 2026 (Article 50 transparency for new systems) · December 2, 2026 (Article 50(2) marking for systems already on market before August 2) Primary source: European Commission, “Guidelines on transparency obligations for providers and deployers of certain AI systems,” July 20, 2026 Source URL: https://digital-strategy.ec.europa.eu/en/policies/guidelines-transparency-ai-generated-content Retrieved: August 1, 2026
July 8, 2026 — Digital Omnibus on AI signed; awaiting Official Journal publication
Type: Legislative milestone — formal adoption complete; OJ published July 24; in force July 27 Status: In force as of July 27, 2026 (see entry above)
The Digital Omnibus on AI completed its full legislative procedure on July 8, 2026, when the Presidents of the European Parliament and the Council of the EU signed the final act. The legislative timeline ran as follows: political agreement May 7, Council Permanent Representatives confirmation May 13, Parliament formal adoption June 16, Council final approval June 29, presidential signing July 8. Publication in the Official Journal of the European Union is expected between July 18 and July 25 — it must occur before July 30 for the regulation to enter into force by August 2, 2026. The Omnibus enters into force on the third day after Official Journal publication. Until that moment, the original AI Act timeline legally remains the operative baseline; however, all major EU law firms and the European Commission are treating the December 2027 Annex III deadline as the operative planning baseline given the near-certain publication schedule. The regulation amends the EU AI Act, civil aviation rules, and the machinery regulation simultaneously.
Compliance deadline affected: Entry into force: 3 days after Official Journal publication (expected late July 2026) · Annex III high-risk: December 2, 2027 · Annex I embedded high-risk: August 2, 2028 · Nudification prohibition: December 2, 2026 · Watermarking: December 2, 2026 Primary source: European Parliament Legislative Observatory; Digital Watch Observatory, “Digital Omnibus on AI: The EU’s AI Act simplification and new AI Office powers,” July 2026 Source URL: https://dig.watch/updates/digital-omnibus-eu-ai-act-new-ai-office-powers Retrieved: July 21, 2026
July 9, 2026 — European Commission issues Opinion on AI-Generated Content Code of Practice
Type: Regulatory action — GPAI / Article 50 transparency Status: Formal Commission opinion published
The European Commission published its formal Opinion assessing the Code of Practice on Transparency of AI-generated content, developed under Article 50 of the EU AI Act. The Opinion marks the institutional endorsement of the transparency labelling framework that will govern chatbots, synthetic media, and AI-generated public-interest content when Article 50 obligations activate on August 2, 2026. The Code covers machine-readable labelling of AI-generated audio, image, video, and text; watermarking obligations for providers of generative AI systems; and disclosure requirements for systems interacting with natural persons. Providers that adhere to the endorsed Code benefit from a compliance presumption under Article 50. Organizations not signed to the Code must demonstrate equivalent compliance through other means from August 2, 2026.
Compliance deadline affected: August 2, 2026 (Article 50 transparency obligations) Primary source: European Commission, Opinion on the assessment of the Code of Practice on Transparency of AI-generated content, July 9, 2026 Source URL: https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai Retrieved: July 13, 2026
July 7, 2026 — Commission publishes EU Action Plan on Cybersecurity and Artificial Intelligence
Type: Policy milestone — AI/cybersecurity intersection Status: Action plan published
The European Commission published its EU Action Plan on Cybersecurity and Artificial Intelligence, establishing a coordinated approach to help member states, businesses, and public authorities address cybersecurity and resilience challenges posed by advanced AI models. The Action Plan includes a Commission commitment to launch a call to increase EU evaluation capacity for AI models before they are placed on the EU market, with full capacity expected by 2027. The Plan intersects with the AI Act’s cybersecurity requirements under Article 15 for high-risk AI systems and the systemic risk assessment obligations for GPAI models with systemic risk. ENISA will co-develop a blueprint for secure access to advanced AI systems for cybersecurity purposes.
Compliance deadline affected: Ongoing — shapes enforcement approach for high-risk AI Article 15 obligations Primary source: European Commission press release, “Commission presents EU Action Plan on Cybersecurity and Artificial Intelligence,” July 7, 2026 Source URL: https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai Retrieved: July 13, 2026
June 16, 2026 — European Parliament formally adopts Digital Omnibus AI amendments (423 votes in favour)
Type: Legislative milestone — binding timeline change Status: Parliament adopted June 16; Council adopted June 29; signed July 8; Official Journal publication pending
The European Parliament voted 423 to 57 (174 abstentions) to formally adopt the Digital Omnibus amendments to the EU AI Act. The vote gives final legislative approval to the timeline changes negotiated in the May 7 trilogue agreement. Key changes now proceeding to Council formal adoption: high-risk AI systems under Annex III have until December 2, 2027 (not August 2, 2026); systems embedded as safety components under Annex I have until August 2, 2028; watermarking obligations under Article 50(2) for AI-generated content are delayed to December 2, 2026; a new prohibition on AI systems that generate non-consensual intimate images (nudification apps) applies from December 2, 2026. The Article 5 prohibitions, GPAI obligations, and penalty ceilings (Article 99) are unchanged. Article 50 transparency obligations — chatbot disclosure, AI-interaction disclosure — activate on August 2, 2026 as originally scheduled.
Compliance deadline affected: Annex III high-risk: December 2, 2027 · Annex I high-risk: August 2, 2028 · Watermarking: December 2, 2026 · Nudification prohibition: December 2, 2026 Primary source: European Parliament press release, “European Parliament approves AI Act amendments, ‘nudifier’ ban,” June 16, 2026 Source URL: https://www.europarl.europa.eu/news/en/press-room/20260427IPR42011/ai-act-deal-on-simplification-measures-ban-on-nudifier-apps Retrieved: July 13, 2026
June 10, 2026 — Commission publishes Code of Practice on AI-generated content marking
Type: Regulatory action — Article 50 quasi-binding guidance Status: Final Code published
The European Commission published the final Code of Practice on marking and labelling of AI-generated content under the EU AI Act. The Code operationalizes Article 50(2) obligations for providers of generative AI systems to ensure that AI-generated audio, image, video, and text content is machine-readable and detectable as artificially generated or manipulated. Signatories benefit from a compliance presumption that they meet the underlying Article 50 requirements. The Code follows the third draft published in March 2026. The first draft was published December 17, 2025; the Commission published its formal endorsement Opinion on July 9, 2026. Providers of generative AI systems serving EU users should assess whether adherence to the Code is their preferred compliance path or whether they will demonstrate compliance through alternative means.
Compliance deadline affected: August 2, 2026 (Article 50 transparency; watermarking extended to December 2, 2026 under Digital Omnibus) Primary source: European Commission, Code of Practice on marking and labelling of AI-generated content, June 10, 2026 Source URL: https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai Retrieved: July 13, 2026
May 7, 2026 — Digital Omnibus trilogue agreement: High-risk deadline shifted to December 2027
Type: Legislative milestone — binding timeline change Status: Political agreement May 7; Parliament adopted June 16; Council adopted June 29; signed July 8; Official Journal publication pending
Parliament and Council negotiators reached a political agreement on the Digital Omnibus on AI — the package of amendments to the EU AI Act proposed by the Commission in November 2025. The agreement: postpones Annex III high-risk AI obligations from August 2, 2026 to December 2, 2027; postpones Annex I embedded high-risk AI obligations to August 2, 2028; adds a new prohibition on AI nudification apps applying from December 2, 2026; extends SME privileges to small mid-cap companies; delays watermarking obligations to December 2, 2026; and strengthens the AI Office’s enforcement powers, including over AI systems built on GPAI models and those embedded in very large online platforms. The Article 5 Article 50 (chatbot disclosure), GPAI obligations (Articles 51–56), and fine ceilings (Article 99) are not affected. The agreement was reached after a failed first attempt, under time pressure to finalize before August 2, 2026.
Compliance deadline affected: Annex III high-risk: December 2, 2027 · Annex I high-risk: August 2, 2028 · Nudification prohibition: December 2, 2026 · Watermarking: December 2, 2026 Primary source: European Commission press release, “EU agrees to simplify AI rules to boost innovation and ban ‘nudification’ apps to protect citizens,” May 7, 2026 Source URL: https://digital-strategy.ec.europa.eu/en/news/eu-agrees-simplify-ai-rules-boost-innovation-and-ban-nudification-apps-protect-citizens Retrieved: July 13, 2026
August 2, 2025 — GPAI model obligations enter application; GPAI Code of Practice published July 10
Type: Enforcement milestone — GPAI obligations operative Status: In force
The rules governing providers of general-purpose AI (GPAI) models under Chapter V of the EU AI Act (Articles 51–56) became applicable on August 2, 2025. From this date, all new GPAI models placed on the EU market must comply with transparency obligations (Article 53), copyright policy requirements, and — for models above the 10²⁵ FLOP systemic risk threshold — safety and security assessment obligations (Article 55). The European AI Office published the final GPAI Code of Practice on July 10, 2025, providing a voluntary compliance tool that creates a presumption of conformity with the underlying legal obligations. Signatories include Anthropic, Google, IBM, Microsoft, Mistral AI, OpenAI, Cohere, and Aleph Alpha. Meta has not signed; xAI signed only the Safety and Security chapter. Chinese AI companies including Alibaba, Baidu, and DeepSeek have not signed. The Commission’s enforcement powers over GPAI providers — including fines of up to 3% of global turnover or €15 million — activate on August 2, 2026. Models placed on the market before August 2, 2025 have until August 2, 2027 to comply.
Compliance deadline affected: GPAI obligations: August 2, 2025 (applicable) · GPAI enforcement powers: August 2, 2026 · Legacy GPAI models: August 2, 2027 Primary source: European Commission, “GPAI Code of Practice,” July 10, 2025; European AI Office, GPAI Signatory Taskforce page Source URL: https://digital-strategy.ec.europa.eu/en/policies/contents-code-gpai Retrieved: July 13, 2026
February 2, 2025 — Article 5 prohibited practices become enforceable across all 27 member states
Type: Enforcement milestone — prohibition regime active Status: In force; investigation priority confirmed
The eight categories of prohibited AI practices under Article 5 of the EU AI Act became enforceable on February 2, 2025. The prohibitions cover: subliminal manipulation techniques that distort persons’ behaviour; exploitation of vulnerabilities of specific groups; public-authority social scoring of natural persons; real-time biometric identification in publicly accessible spaces by law enforcement (with limited exceptions); biometric categorization to infer sensitive attributes; emotion recognition in workplaces and educational institutions; individual criminal risk assessment based solely on profiling; and — following the Digital Omnibus agreement — AI systems designed to generate non-consensual intimate images (effective December 2, 2026). Penalties for prohibited practice violations reach €35 million or 7% of global annual turnover, whichever is higher. AI literacy obligations for providers and deployers entered application on the same date. As of the tracker update date, no Article 5 fine has been publicly confirmed by any national competent authority or the European AI Office; multiple investigations are reported underway in workplace emotion recognition and social scoring contexts.
Compliance deadline affected: Article 5 prohibited practices: February 2, 2025 (already in force) · Penalty exposure for ongoing violations accumulates from this date Primary source: Regulation (EU) 2024/1689, Article 5; European Commission, AI Act Implementation FAQ Source URL: https://digital-strategy.ec.europa.eu/en/faqs/navigating-ai-act Retrieved: July 13, 2026
August 1, 2024 — EU AI Act enters into force
Type: Legislative milestone — foundational Status: In force
Regulation (EU) 2024/1689 entered into force on August 1, 2024, one day after publication in the Official Journal of the European Union on July 12, 2024. The Act is the world’s first comprehensive legal framework on artificial intelligence. Its application follows a phased schedule: Article 5 prohibitions and AI literacy obligations applied from February 2, 2025; GPAI model obligations and governance provisions applied from August 2, 2025; the main body of the Act — transparency obligations (Article 50), market surveillance, penalty framework, and GPAI enforcement powers — applies from August 2, 2026; Annex III high-risk AI obligations now apply from December 2, 2027 following the Digital Omnibus; Annex I embedded high-risk AI applies from August 2, 2028; and legacy public-sector high-risk AI systems must comply by August 2, 2030.
Compliance deadline affected: Foundation date for all subsequent enforcement deadlines Primary source: Regulation (EU) 2024/1689, Official Journal of the European Union, July 12, 2024 Source URL: https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai Retrieved: July 13, 2026
Axis Intelligence Original Analysis: The Enforcement Gap Ratio™
Axis Intelligence Research tracks one figure that no EU institution publishes directly: the Axis Intelligence EU AI Act Enforcement Gap Ratio™ — the share of EU member states that had formally notified at least one single point of contact (SPOC) to the European Commission as required under Article 70, measured against the total of 27 member states, at each tracker update.
As of July 13, 2026: Based on the European Parliament Research Service figure cited in published legal analyses (regdossier.eu, April 2026), 8 of 27 member states had notified their SPOC as of March 2026. This represents a 29.6% notification rate at the last confirmed reporting point. Axis Intelligence Research will update this figure quarterly when the Commission’s annual enforcement report is published per Article 99(11). Until then, the March 2026 figure is the most recent verified data point.
Formula: (Member states with confirmed SPOC notification ÷ 27) × 100 = Enforcement Gap Ratio Inputs: European Parliament Research Service report cited in regdossier.eu analysis, April 2026; Article 70 and Article 99(11) of Regulation (EU) 2024/1689 Limitation: This figure reflects formal SPOC notification only, not operational enforcement capacity. A notified SPOC does not imply a fully staffed, investigation-ready national competent authority. The ratio understates the practical enforcement gap.
Corroborating data point (added August 24, 2026): A separate mid-2026 assessment reported by informedclearly.com found that 10 of 27 member states — Ireland, Spain, Lithuania, Finland, France, Germany, Netherlands, Poland, Cyprus, and Italy — showed advanced public implementation evidence, with Ireland designating 15 competent authorities and Spain’s AESIA publishing 16 practical compliance guides. Seventeen states had limited public footprints. Axis Intelligence Research does not merge this figure with the SPOC Enforcement Gap Ratio: the two measure different things (formal notification versus observable implementation activity) and combining them would produce a number that means nothing. Both are published side by side. Cullen International’s April 2026 survey separately found that Belgium, the Czech Republic, Finland, Germany, Ireland, Luxembourg, the Netherlands, Romania, and Sweden assigned enforcement powers to their telecoms regulators, and that only Denmark, Finland, and Italy had a national AI law in place.
Why this matters for enforcement: The uneven national picture is the structural reason the first enforcement wave is likely to come from the AI Office over GPAI models rather than from national market surveillance authorities over deployed AI systems. The AI Office is a single, centrally resourced body currently recruiting 40 additional contractual agents. Twenty-seven national authorities at widely varying readiness levels are not.
Axis Intelligence EU AI Act Enforcement Timeline — Verified Summary
| Obligation | Applies from | Penalty ceiling | Enforcer | Status as of August 24, 2026 |
|---|---|---|---|---|
| Article 5 prohibited practices | Feb 2, 2025 | €35M or 7% turnover | National competent authorities | In force; no public fine confirmed |
| AI literacy (Art. 4) | Feb 2, 2025 | Via Article 99 | National competent authorities | In force |
| GPAI model obligations (Arts. 51–56) | Aug 2, 2025 | €15M or 3% turnover | European AI Office | In force |
| Digital Omnibus (Reg. EU 2026/1744) | July 27, 2026 | N/A | N/A | In force |
| Article 50 transparency (chatbot, deepfake, AI content) | Aug 2, 2026 | Via Article 99 | National competent authorities + AI Office | Active 22 days — no formal decision published |
| GPAI enforcement powers (Art. 101) | Aug 2, 2026 | €15M or 3% turnover | European AI Office | Active 22 days — dialogues underway; no formal decision published |
| Market surveillance & full penalty framework | Aug 2, 2026 | €35M / €15M / €7.5M tiers | National competent authorities | Active 22 days — no formal decision published |
| Watermarking — pre-existing systems (Art. 50(2)) | Dec 2, 2026 | Via Article 99 | National competent authorities | Grace period active for systems on market before Aug 2, 2026 |
| AI nudification app prohibition | Dec 2, 2026 | €35M or 7% turnover | National competent authorities | Binding law; applies December 2, 2026 |
| Annex III high-risk stand-alone AI | Dec 2, 2027 | €15M or 3% turnover | National competent authorities | Delayed — binding law since July 27, 2026 |
| Legacy GPAI models (pre-Aug 2025) | Aug 2, 2027 | €15M or 3% turnover | European AI Office | In preparation |
| Annex I embedded high-risk AI | Aug 2, 2028 | €15M or 3% turnover | National competent authorities | Delayed — binding law since July 27, 2026 |
Source: Regulation (EU) 2024/1689; Regulation (EU) 2026/1744 (OJ July 24, in force July 27, 2026); European Commission enforcement framework page; Commission press release IP/26/1714. Axis Intelligence Research compilation, August 24, 2026.
Frequently Asked Questions
Has any company been fined under the EU AI Act yet?
As of August 24, 2026 — 22 days into the enforcement era — no fine, formal investigation opened by Commission decision, or market withdrawal order under the EU AI Act has been publicly confirmed by the European AI Office, the European Data Protection Supervisor, or any national competent authority. Axis Intelligence Research verified this across the Commission presscorner, the AI Office page, the enforcement framework page, and the AI Act Service Desk on August 24, 2026. The AI Office has stated that technical compliance dialogues are its preferred initial tool before formal escalation. Notably, the Commission has confirmed that GPAI violations dating back to August 2, 2025 fall within enforcement scope — the twelve-month enforcement gap does not shield earlier non-compliance. This tracker adds an entry within 48 hours of any formally confirmed enforcement action carrying a case number or official Commission press release.
What does the Digital Omnibus actually change?
The Digital Omnibus is now binding law — Regulation (EU) 2026/1744, published in the Official Journal July 24, in force July 27, 2026. Three things changed; three did not. Changed: Annex III high-risk AI obligations moved to December 2, 2027; Annex I embedded high-risk AI moved to August 2, 2028; watermarking for pre-existing systems (Article 50(2)) moved to December 2, 2026; a new prohibition on nudification apps applies from December 2, 2026. Unchanged: Article 5 prohibitions, GPAI model obligations, Article 50 chatbot and AI-interaction transparency (which activate August 2, 2026), and the penalty ceilings under Article 99.
Does the EU AI Act apply to US companies?
Yes, extraterritorially. Any provider or deployer whose AI output is used in the EU is in scope regardless of where the organization is headquartered. A US company that sells software with AI features to EU customers or deploys AI affecting persons in the EU must classify its systems and comply with the applicable obligations. The “Brussels Effect” operates through market access: non-compliance means market withdrawal, not just a fine.
What is the GPAI Code of Practice and who has signed it?
The GPAI Code of Practice is a voluntary compliance tool published by the European AI Office on July 10, 2025 under Article 56 of the EU AI Act. Signatories can demonstrate compliance with their GPAI model obligations (Articles 53 and 55) by adhering to the Code. As of late July 2026, approximately 190 organisations had signed the Code, per the European Commission’s Code of Practice page. The Article 50 Transparency Code of Practice — a separate instrument — had more than 180 signatories as of July 31, 2026, per the Commission’s enforcement press release. Confirmed GPAI Code signatories include Anthropic, Google, IBM, Microsoft, Mistral AI, OpenAI, Cohere, and Aleph Alpha. Meta has not signed the GPAI Code. xAI signed only the Safety and Security chapter.
What happens on August 2, 2026?
Three enforcement mechanisms activate simultaneously: (1) The European Commission gains authority to impose fines on GPAI model providers under Article 101 — up to €15 million or 3% of global turnover; (2) Article 50 transparency obligations become fully enforceable, requiring chatbot disclosure, AI-interaction notification, and deepfake labelling for new systems — pre-existing systems placed on market before August 2 have a grace period until December 2, 2026 for the machine-readable marking obligation; (3) National competent authorities across all 27 member states gain full market surveillance powers, including the right to investigate, order withdrawals, and levy fines. Annex III high-risk AI obligations do not activate on August 2 — Regulation (EU) 2026/1744 (in force July 27) moves that deadline to December 2, 2027. The Commission formally announced enforcement commencement in a press release dated July 31, 2026 and published over 180 Article 50 Code of Practice signatories on the same date.
What is the penalty ceiling for EU AI Act violations?
The Act uses a three-tier structure: €35 million or 7% of global annual turnover (whichever is higher) for Article 5 prohibited practices; €15 million or 3% for other infringements including high-risk AI and GPAI obligations; €7.5 million or 1.5% for providing incorrect or misleading information to authorities. For SMEs and startups, Article 99(6) specifies the lower of the fixed amount and the percentage-of-turnover threshold applies. These ceilings apply per infringement, not per investigation. The penalty structure does not change under the Digital Omnibus.
How Often This Tracker Updates
Axis Intelligence Research reviews this tracker weekly. New entries are added within 48 hours of a primary-source document becoming public from the European Commission, European AI Office, European Parliament, Council of the EU, or a national competent authority. The tracker updates on the following trigger events regardless of the weekly schedule: any publicly confirmed enforcement action (fine, market withdrawal, formal investigation) from any EU institution or national authority; any publication from the AI Office affecting the GPAI Code of Practice; any formal change to the EU AI Act application timeline.
Review trigger: Next scheduled update August 30, 2026. Unscheduled update within 48 hours of: any publicly confirmed EU AI Act enforcement action — investigation opened by formal decision, fine imposed, or market withdrawal ordered — carrying a case number or official Commission press release from the European AI Office, the European Data Protection Supervisor, or a national competent authority.
Download the Dataset
The entries in this tracker are available as a downloadable dataset in CSV format, licensed CC BY 4.0. The CSV is the fact table: every entry in the article has a corresponding row with source provenance columns. Cite as: Axis Intelligence Research, EU AI Act Compliance Tracker, July 2026, axis-intelligence.com/eu-ai-act-compliance-tracker/.
Download: eu-ai-act-compliance-tracker.csv
Citation Block
APA: Axis Intelligence Research. (2026, July 13). EU AI Act compliance tracker 2026 — every milestone, live status. Axis Intelligence. https://axis-intelligence.com/eu-ai-act-compliance-tracker/
MLA: Axis Intelligence Research. “EU AI Act Compliance Tracker 2026 — Every Milestone, Live Status.” Axis Intelligence, 13 July 2026, axis-intelligence.com/eu-ai-act-compliance-tracker/.
Chicago: Axis Intelligence Research. “EU AI Act Compliance Tracker 2026 — Every Milestone, Live Status.” Axis Intelligence, July 13, 2026. https://axis-intelligence.com/eu-ai-act-compliance-tracker/.
Go Deeper — Axis Intelligence Research
- EU AI Act Compliance Checker — assess your organization’s risk tier and obligations in under two minutes, updated post-Digital Omnibus
- EU AI Act Statistics 2026 — the regulation in numbers: adoption timelines, enforcement readiness, and compliance cost benchmarks
- EU AI Act Fines Statistics — penalty ceiling analysis and the methodology behind the three-tier fine structure
- EU AI Act Compliance Cost Statistics — what readiness actually costs, by organization size and risk tier
- EU AI Act Compliance Statistics — readiness levels across EU industries and reported compliance gaps
- EU AI Act News 2026 — Digital Omnibus updates, enforcement notices, and national authority guidance tracked as it lands
Authored by Axis Intelligence Research and Sarah Mitchell, AI/ML Editor. All entries sourced from primary EU institutional documents fetched and verified at entry time. Axis Intelligence Research does not accept retainers, consulting fees, or compensation from any organization whose compliance status is tracked in this report. Not legal advice. Consult qualified legal counsel for compliance decisions. 13, 2026. https://axis-intelligence.com/eu-ai-act-compliance-tracker/.