EU AI Act Amendments Agreed: New Prohibitions, Enhanced Oversight, and SME Support
Political agreement reached on AI Act amendments including bans on non-consensual intimate content, reinforced AI Office powers, and extended SME relief measures.
Major Amendments Agreed
A political agreement on amendments to the EU AI Act was reached on 7 May 2026, following the proposal’s adoption on 19 November 2025. The amendments introduce significant new protections and governance clarifications across the bloc.
New Prohibited Practices
The amendments include a prohibition of AI systems that generate non-consensual sexually explicit and intimate content or child sexual abuse material, such as AI ‘nudification’ apps.
This adds to the eight practices already prohibited under the AI Act: harmful AI-based manipulation and deception, harmful AI-based exploitation of vulnerabilities, social scoring, individual criminal offence risk assessment or prediction, untargeted scraping of the internet or CCTV material to create or expand facial recognition databases, emotion recognition in workplaces and education institutions, biometric categorisation to deduce certain protected characteristics, and real-time remote biometric identification for law enforcement purposes in publicly accessible spaces.
Strengthened Governance
The AI Office’s powers have been reinforced and oversight of AI systems built on general-purpose AI models has been centralised to reduce governance fragmentation.
Extended SME Relief
Simplified requirements previously granted to small and medium-sized enterprises (SMEs) have been extended to small mid-cap companies (SMCs), including simplified technical documentation requirements.
Expanded Sandbox Access
More innovators will gain access to regulatory sandboxes, including an EU-level sandbox, to test AI solutions in real-world conditions.
Regulatory Clarity
The interplay between the AI Act and EU product safety laws, in particular the Machinery Regulation, has been clarified to avoid duplication between sectoral and AI rules.
Timeline of Implementation
The AI Act entered into force on 1 August 2024 and will be fully applicable on 2 August 2026. Key milestones include:
- 2 February 2025: Prohibited AI practices and AI literacy obligations entered into application
- 2 August 2025: Governance rules and obligations for GPAI models became applicable
- 2 December 2027: Rules for high-risk AI systems used in biometrics, critical infrastructure, education, employment, migration, asylum and border control will apply
- 2 August 2028: Rules for high-risk AI systems embedded into regulated products and integrated into products such as lifts or toys will apply
Ireland’s Implementation Path
The General Scheme of the Regulation of Artificial Intelligence Bill 2026 was published on 4 February 2026 and last updated on 9 February 2026. This gives further effect to Government Decisions of 4 March 2025 and 22 July 2025 approving that Ireland will adopt a distributed model of competent authorities for the AI Act.
The General Scheme proposes to establish a new statutory independent body – Oifig Intleachta Shaorga na hÉireann, the AI Office of Ireland, under the remit of the Department of Enterprise, Tourism and Employment, to act as the Single Point of Contact and central coordinating authority for the implementation and enforcement of the EU AI Act in Ireland.
Developments since publication
-
The Irish Regulation of Artificial Intelligence Bill completed the Seanad Second Stage on 1 July 2026 and is currently in the Seanad Committee Stage Source
-
The UK's AI Regulation and Safety Bill passed its second Lords reading on July 3, 2026 Source
-
China began enforcing companion AI rules on July 15, 2026 Source
-
On August 2, 2026, the European Commission can start fining general-purpose AI providers, and Article 50 transparency rules switch on across all 27 member states Source
-
The European Commission presented an Action Plan on Cybersecurity and Artificial Intelligence on 7 July 2026, developed with the European Union Agency for Cybersecurity (ENISA) Source
-
The European Commission published guidelines on transparency obligations for providers and deployers of certain AI systems on 20 July 2026 Source
-
Threat actors are exploiting CVE-2026-6875, a sandbox escape vulnerability in ServiceNow AI Platform that allows unauthenticated users to run arbitrary code. Source
-
CVE-2026-6875 has a CVSS score of 9.5. Source
-
ServiceNow released patches for CVE-2026-6875 throughout June in versions: Brazil EA/GA, Australia Patch 2, Zurich Patch 7b/9, and Yokohama Patch 12 Hot Fix 1b/Patch 13. Source
-
Searchlight Cyber reported the CVE-2026-6875 vulnerability to ServiceNow on April 1, 2026. Source
-
The CVE-2026-6875 exploitation targets the pre-authentication endpoint /assessment_thanks.do using HTTP POST requests. Source
-
Attackers are achieving code execution through a different sandbox-escape gadget chain than the one documented in public proof-of-concept exploits. Source
-
81.2% of companies running AI packages have at least one known vulnerability. Source
-
99.9% of AI vulnerability alerts with an available fix remain unpatched. Source
-
74.1% of companies running AI packages have at least one critical CVE. Source
-
56% of AI adopters have deployed agent frameworks into production. Source
-
51.5% of AI adopters use AI to build custom applications. Source
-
Between 87% and 98% of organizations across the three major cloud providers have not configured customer-managed encryption keys for their AI services. Source
-
64% of AI adopters have deployed vector databases that connect LLMs to internal documents, customer records, and proprietary knowledge. Source
-
Businesses using retrieval-augmented generation operate an average of 3.78 vector databases. Source
-
Nearly 30% of AI adopters store at least one AI key in an insecure location. Source
-
The EU AI Act introduces additional requirements for high-risk AI systems beginning on August 2, 2026. Source
-
Colorado's amended AI law takes effect on January 1, 2027. Source
-
China has expanded its cybersecurity framework with AI-specific requirements and mandatory labeling of AI-generated content. Source
-
ServiceNow is enhancing instance security by severely restricting the type of code that can run in sandbox contexts. Source
-
The Digital Package on Simplification (AI omnibus) has been adopted on 19 November 2025 and a political agreement was reached on 7 May 2026. Source
-
The final text of the AI Omnibus Regulation entered into force in July 2026. Source
-
Ireland's Regulation of Artificial Intelligence Bill 2026 establishes Oifig IS na hÉireann (AI Office of Ireland) as an independent statutory body which will act as Ireland's central coordinating auth Source
-
The EU AI Act entered into force in August 2024, with its provisions applying in a phased manner, and applies directly across all EU Member States. Source
-
Ireland's Regulation of Artificial Intelligence Bill 2026 is a technical implementing measure and does not add to the obligations placed on regulated entities by the EU Regulation. Source
-
The Bill is structured in 10 Parts with 139 Sections and 4 Schedules. Source
-
Ireland's Bill empowers Market Surveillance Authorities (MSAs) with a structured, proportionate enforcement toolkit, progressing from cooperative compliance notices to coercive measures such as prohib Source
-
The European Union published the Cloud and AI Development Act in the Official Journal of the EU on July 15, 2026. Source
-
The Cloud and AI Development Act applies broadly to any company offering cloud services—including Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS)—as Source
-
Ireland established 15 specialised enforcement authorities for the EU AI Act. Source
-
The European Commission published the final version of its Code of Practice on Transparency of AI-Generated Content. Source
-
The core transparency obligations under Article 50 of the AI Act take effect on 2 August 2026. Source
-
Ireland's Regulation of Artificial Intelligence Bill 2026 was published on 17 June 2026. Source
-
Ireland's Regulation of Artificial Intelligence Bill 2026 is designed to give full domestic effect to the EU Artificial Intelligence Act. Source
-
Prohibited AI practices including harmful AI-based manipulation and deception, harmful AI-based exploitation of vulnerabilities, and social scoring became effective in February 2025. Source
-
A political agreement on the Digital Omnibus simplification amendments to the AI Act was reached on 7 May 2026. Source
-
The AI Act amendments reinforce the AI Office's powers and centralise oversight of AI systems built on general-purpose AI models, reducing governance fragmentation. Source
-
Minister Niamh Smyth launched a new industry consultation inviting businesses from across the manufacturing, tourism, construction and ICT services sectors to help shape the future of AI adoption in I Source
-
Accenture has been commissioned by the Department of Enterprise, Tourism and Employment to undertake a study examining opportunities and barriers to AI adoption in the manufacturing, tourism and const Source
-
The consultation for Ireland's AI in Enterprise study will close on 24 July 2026. Source
-
The submissions from Ireland's AI in Enterprise consultation will inform the first phase of the department's Sectoral Strategy for Enterprise Adoption of AI. Source
-
The Digital Omnibus on AI was signed on 8 July 2026 and is awaiting publication in the Official Journal. Source
-
Stand-alone Annex III high-risk systems now have until 2 December 2027 for compliance, extended from the original 2 August 2026 deadline. Source
-
AI embedded in regulated products under Annex I has until 2 August 2028 for compliance, extended from the original 2 August 2026 deadline. Source
-
A new prohibition on AI-generated non-consensual intimate imagery was added to Article 5 of the EU AI Act. Source
-
The AI Office gained broader supervisory reach over vertically integrated AI providers under the Digital Omnibus amendments. Source
-
Systems that interact directly with people must make clear that a person is dealing with a machine, unless that is obvious from context, applying from 2 August 2026. Source
-
Deepfakes and manipulated audio, image or video content must be disclosed as artificially generated, applying from 2 August 2026. Source
-
A political agreement was reached on 7 May 2026 on the Digital Package on Simplification, dubbed the AI omnibus, proposed on 19 November 2025. Source
-
The Commission published draft guidelines to clarify how the transparency obligations under Article 50 should be applied in practice on 8 May 2026. Source
-
The rules for high-risk AI systems embedded into regulated products have an extended transition period until 2 August 2028 as a result of the political agreement on the proposal to simplify the AI Act Source
-
The AI Act entered into force on 1 August 2024 and will be fully applicable 2 years later on 2 August 2026, with exceptions for prohibited AI practices, AI literacy obligations (2 February 2025), GPAI Source
-
The political agreement on the AI omnibus simplification proposal was reached on 7 May 2026, setting clear timelines: rules for high-risk systems in certain areas apply from 2 December 2027, and rules Source
-
A new AI Office of Ireland will be established as a central and coordinating authority by August 2026 for the implementation of the AI Act in Ireland Source
-
Ireland adopted a distributed model of implementation of the EU AI Act, building on the capacity and expertise of established sectoral regulators rather than a single centralized authority Source
-
Ireland has designated 15 national competent authorities for the EU AI Act implementation Source
-
For infringements of prohibited AI practices or non-compliance with data requirements, maximum penalties are up to €35 million or 7% of total worldwide annual turnover (whichever is higher) Source
-
Ireland's fundamental rights authorities will have additional powers from 2 August 2026 to access documentation that developers and deployers of AI systems are required to hold under the AI Act Source
-
Rules on prohibited AI practices became effective in February 2025 Source
-
Rules on high-risk AI system use cases (Annex III) come into effect on 2 August 2026 Source
-
The European Commission presented the EU Action Plan on Cybersecurity and Artificial Intelligence on July 7, 2026, to strengthen enforcement of the AI Act by establishing dedicated evaluation capacity Source
-
The July 2026 EU Action Plan on Cybersecurity and AI sets out a coordinated approach to help Member States, businesses and public authorities address cybersecurity and resilience challenges posed by t Source
-
The Commission will launch a call to increase EU evaluation capacity of AI models before they are placed in the EU market, expected to be operational by 2027. Source
-
The new EU evaluation capacity will strengthen third-party assessment of AI capabilities and risks and contribute to the regulatory function of the AI Office. Source
-
The Commission and the European Union Agency for Cybersecurity (ENISA) will create a blueprint to secure access to advanced AI systems for cybersecurity purposes and establish a secure testing platfor Source
-
Enforcement of the AI Act's key provisions for advanced models begins on August 2, 2026, requiring developers to demonstrate risk evaluations and mitigation strategies. Source
-
The plan does not introduce new legal requirements but enhances the ability to meet existing AI Act obligations through EU-supported evaluation resources. Source
-
The plan builds explicitly on AI Act provisions for general-purpose AI models with systemic risks, including the GPAI Code of Practice. Source
-
Simplified requirements granted to small and medium-sized enterprises under the AI Act are being extended to small mid-cap companies (up to 750 employees and €150 million annual revenue). Source
-
The General Scheme provides rules on penalties for infringement of the AI Act in Ireland. Source
-
Guidelines on transparent AI systems will clarify the scope of application, relevant legal definitions, the transparency obligations, the exceptions and related horizontal issues. Source
-
The prohibitions on unacceptable risk AI practices became effective in February 2025. Source
-
The AI Act rules on General-Purpose AI became effective in August 2025. Source
-
Following the political agreement on the proposal to simplify the AI Act, for systems integrated into products such as lifts or toys, the rules will apply from 2 August 2028. Source
-
The AI Act is used by more than 150,000 users every month. Source
-
According to Article 57 of the AI Act, each Member State must establish at least one AI regulatory sandbox at the national level by 2 August 2026. Source
-
The AI Act entered into force on 1 August 2024 and will be fully applicable 2 years later on 2 August 2026, with some exceptions including prohibited AI practices and AI literacy obligations that ente Source
-
High-risk AI systems covered by the AI Omnibus Agreement face compliance deadlines of 2 December 2027 for systems in Annex III (covering employment, education, and health insurance) and 2 August 2028 Source
-
AI systems that generate or manipulate non-consensual sexually explicit and intimate content or child sexual abuse material are prohibited, with the prohibition taking effect from 2 December 2026. Source
-
The AI Office's powers are reinforced to centralise oversight of AI systems built on general-purpose AI models, reducing governance fragmentation. Source
-
Ireland will adopt a distributed model of competent authorities for the AI Act, leveraging established sectoral regulatory authorities, with a designated central authority to provide coordination and Source
-
The General Scheme of the Regulation of Artificial Intelligence Bill 2026 gives further effect to Government Decisions of 4 March 2025 and 22 July 2025. Source
-
Prohibited AI practices under the AI Act became effective in February 2025. Source
-
Generative AI systems placed on the market or put into service before 2 August 2026 must comply with watermarking requirements only as of 2 December 2026. Source
-
AI systems designed to create intimate deepfakes or CSAM are prohibited effective 2 December 2026, with violations triggering fines of up to €35 million or 7% of annual worldwide turnover, whichever i Source
-
On 7 May 2026, EU legislative bodies reached a political agreement on proposed amendments to the AI Act, known as the AI Act Omnibus. Source
-
The EU AI Act prohibits the use of GDPR special category personal data for bias detection purposes except where necessary to detect and mitigate bias in AI models. Source
-
Formal adoption and publication of the AI Act amendments in the Official Journal of the European Union were anticipated in July 2026, ahead of the 2 August 2026 deadline. Source
-
Fable 5 returns globally on July 1. Source
-
Anthropic is proposing an industry-wide framework for scoring jailbreak severity together with Amazon, Microsoft, Google, and other Glasswing partners. Source
-
Anthropic launched Claude Sonnet 5 delivering frontier performance across coding, agents, and professional work at scale. Source
-
Claude Science, a customizable AI workbench for scientists, integrates tools and packages researchers most often use, produces auditable artifacts, and provides flexible access to computing resources. Source
-
A political agreement on the AI omnibus was reached on 7 May 2026. Source
-
The AI Act will be fully applicable on 2 August 2026. Source
-
On 29 June, the Council of the EU gave its final green light to the AI Act simplification package, following the European Parliament's formal endorsement on 16 June 2026. Source
-
The legislative act will be published in the EU's official journal shortly and will enter into force on the third day after this publication. Source
-
On 16 June, the European Parliament formally endorsed the provisional agreement reached in May regarding the AI Act simplification package. Source
-
The Act classifies AI systems used in employment-related decisions as high-risk, including tools used for recruitment, candidate selection, performance evaluation, task allocation, monitoring of worke Source
-
The European Commission published the Digital Omnibus on AI on 19 November 2025, proposing to defer the high-risk compliance deadline from 2 August 2026 to 2 December 2027. Source
-
The Act's timely implementation has faced significant delays, particularly around the designation of national competent authorities and the finalisation of harmonised standards and compliance tools ne Source
-
Prohibited AI practices — including harmful AI-based manipulation, social scoring, individual criminal offence risk assessment, untargeted biometric scraping, emotion recognition in workplaces, biomet Source
-
A political agreement on amendments to the AI Act was reached on 7 May 2026, introducing prohibition of AI systems that generate non-consensual sexually explicit and intimate content or child sexual a Source
-
The AI Act amendments extend simplified requirements granted to small and medium-sized enterprises (SMEs) to small mid-cap companies (SMCs). Source
-
The Commission will clarify the interplay between the AI Act and EU product safety laws, in particular the Machinery Regulation, to avoid duplication between sectoral and AI rules. Source
-
Guidelines on transparent AI systems will be published in the second quarter of 2026 to clarify transparency obligations under the AI Act. Source
-
Certain simplified requirements granted to small and medium-sized enterprises are extended to small mid-cap companies in agreed amendments to the AI Act. Source
-
In July 2025, the Commission published 3 key instruments to support the responsible development and deployment of GPAI models: Guidelines on the scope of the obligations for providers of GPAI models, Source
-
The Commission published guidelines on prohibited AI practices under the AI Act to offer legal explanations and practical examples to help stakeholders understand and comply with the prohibitions. Source
-
The Commission published guidelines on the AI system definition of the AI Act to assist stakeholders in determining the scope of the AI Act. Source
-
Certain simplified requirements granted to small and medium-sized enterprises are extended to small mid-cap companies (SMEs and SMCs), including simplified technical documentation requirements. Source
-
The Commission publishes Code of Practice on marking and labelling of AI-generated content on 10 June 2026. Source
-
Commission selects EUROPA consortium as the winner of the Frontier AI Grand Challenge, a project to build European open-source frontier AI model in all 24 EU languages, announced 19 June 2026. Source
-
On 7 May 2026, negotiators from the Council of the European Union, the European Parliament, and the European Commission reached a provisional agreement on the terms of the Digital Omnibus on AI, marki Source
-
For AI systems generating or manipulating synthetic content placed on the EU market or put into service before 2 August 2026, the provider's obligation to ensure the system's outputs are marked as art Source
-
Member States are now required to establish at least one regulatory sandbox at the national level by 2 August 2027 instead of 2 August 2026 (deferred by 1 year). Source
-
The new prohibitions on non-consensual intimate material and child sexual abuse material generated by AI systems take effect on 2 December 2026. Source
-
AI literacy requirements in the amended Act now require providers and deployers to take measures to support the development of AI literacy among staff instead of requiring them to ensure AI literacy. Source
-
The amendments are expected to proceed through formal adoption with final approval anticipated in June 2026 and publication expected in July 2026. Source
-
MIT researchers found that a small AI model can outperform the biggest ones at 1 percent of the cost when used as a test bed for AI agents in the game Battleship. Source
-
MIT created the ChartNet training dataset to improve the accuracy of vision-language models that help analyze business trends or interpret scientific figures. Source
-
The prohibitions on eight harmful AI practices became effective in February 2025. Source
-
Simplified technical documentation requirements granted to small and medium-sized enterprises have been extended to small mid-cap companies. Source
-
On May 7, 2026, EU legislative bodies (the Council, European Parliament, and European Commission) reached a political agreement on proposed amendments to the AI Act called the 'AI Act Omnibus'. Source
-
The agreement extends compliance deadlines for stand-alone high-risk AI systems to December 2, 2027, a deferral of 16 months from the original August 2, 2026 deadline. Source
-
For AI systems that are regulated products or safety components (such as medical devices, lifts, radio equipment), the compliance deadline is extended to August 2, 2028, a deferral of 1 year from the Source
-
Effective December 2, 2026, the agreement extends prohibitions to 'nudifier' applications—AI systems that generate or manipulate sexually explicit or intimate images, video, or audio without explicit Source
-
Violations of the prohibition on nudifier applications and CSAM may trigger fines of up to €35 million or 7% of annual worldwide turnover, whichever is higher. Source
-
For AI systems generating or manipulating synthetic content placed on the EU market before August 2, 2026, the watermarking requirement deadline is extended from August 2, 2026 to December 2, 2026 (a Source
-
Violations of watermarking transparency requirements may result in fines of up to €15 million or 3% of total annual worldwide turnover, whichever is higher. Source
-
The AI Act's simplified compliance framework for small and medium-sized enterprises is extended to companies with up to 750 employees and €150 million in annual revenue. Source
-
AI used in industrial applications and products already regulated under the Machinery Regulation is exempt from the AI Act. Source
-
The amendments make it easier to use GDPR special category personal data (health information, biometric data, race, or sexual orientation) where necessary to detect and mitigate bias in AI models. Source
-
The text of the provisional agreement will proceed to formal adoption by the European Parliament and the Council, expected by July 2026. Source
-
Prohibited AI practices, including social scoring, subliminal manipulation, and real-time biometric remote identification in public spaces, became effective in February 2025. Source
-
The AI Act's Code of Practice on marking and labelling of AI-generated content and Guidelines on transparent AI systems will be published in the second quarter of 2026. Source
-
The AI Act entered into force on 1 August 2024, with full applicability set for 2 August 2026, with certain exceptions for prohibited practices and GPAI rules. Source
-
Obligations for high-risk AI systems placed on the market or put into service before 2 August 2026 will be postponed from 2 August 2026 to 2 December 2027 (deferred by 16 months) for use-based applica Source
-
Transparency obligations for synthetic content detection (Article 50) are deferred from 2 August 2026 to 2 December 2026 for systems placed on market before 2 August 2026; systems placed after 2 Augus Source
-
The AI Act establishes simplified requirements for small and medium-sized enterprises (SMEs) that are being extended to small mid-cap companies (SMCs) as part of the Digital Omnibus on AI. Source
-
The AI Office's powers are being reinforced and oversight of AI systems built on general-purpose AI models is being centralised to reduce governance fragmentation as part of the Digital Omnibus on AI. Source
-
The Commission made €63.2 million available to support AI innovation in health and online safety on 21 April 2026. Source
-
The Code of Practice on marking and labeling AI-generated content is under preparation and scheduled for publication in the second quarter of 2026. Source
-
AI systems that generate non-consensual sexually explicit and intimate content or child sexual abuse material are prohibited Source
-
A political agreement on amendments to simplify the AI Act was reached on 7 May 2026 Source
-
The AI Act entered into force on 1 August 2024 Source
-
The Code of Practice on marking and labelling of AI-generated content is a voluntary tool to guide providers and deployers of generative AI systems on transparency obligations including marking AI gen Source
-
The US government issued an emergency export control directive on June 12, 2026, ordering Anthropic to immediately suspend all access to Fable 5 and Mythos 5 for any foreign national Source
-
The EU has published its final Code of Practice on AI-generated content transparency with an August 2 enforcement date, and OpenAI has stated its commitment to sign Source
-
The European AI Office will publish the first public list of Code of Practice signatories in July 2026 Source
-
The EU Cyber Resilience Act's conformity assessment body notification framework entered into application on June 11, 2026, starting a 90-day window before September 11, when mandatory vulnerability an Source
-
NIST has provided a mathematical basis using Gödel's incompleteness logic showing that a fixed set of AI guardrails cannot theoretically prevent all adversarial bypasses Source
-
The AI Act is the first-ever comprehensive legal framework on AI worldwide. Source
-
Eight AI practices are prohibited under the AI Act, including harmful manipulation, exploitation of vulnerabilities, social scoring, criminal risk prediction, untargeted scraping for facial recognitio Source
-
Prohibitions on AI practices became effective in February 2025. Source
-
The transparency rules of the AI Act will come into effect in August 2026. Source
-
High-risk AI systems rules for systems used in certain high-risk areas will apply from 2 December 2027. Source
-
A prohibition on AI systems that generate non-consensual sexually explicit and intimate content or child sexual abuse material takes effect on 2 December 2026. Source
-
On 7 May 2026, negotiators from the Council of the European Union, European Parliament, and European Commission reached a provisional agreement on the Digital Omnibus on AI. Source
-
Obligations for Annex III high-risk AI systems (use-based) are postponed from 2 August 2026 to 2 December 2027, a deferral of 16 months. Source
-
Obligations for Annex I high-risk AI systems (product-regulated, including radio equipment, lifts and medical devices) are postponed from 2 August 2027 to 2 August 2028, a deferral of 1 year. Source
-
The obligation for Member States to establish at least one national AI regulatory sandbox is postponed from 2 August 2026 to 2 August 2027. Source
-
For AI systems generating or manipulating synthetic content placed on the EU market before 2 August 2026, the obligation to ensure outputs are marked in machine-readable format and detectable as artif Source
-
The Machinery Regulation has been moved from Annex I Section A to Section B of the AI Act, shifting AI-enabled machinery from a dual-compliance model to one where sector-specific laws are paramount. Source
-
AI systems used solely for user assistance, performance optimization, efficiency, automation, convenience or quality control will not qualify as safety components under Article 6(1) unless a failure c Source
-
The prohibition on AI-generated non-consensual intimate material and CSAM applies to providers where such generation is the system's intended purpose or where it is a reasonably foreseeable and reprod Source
-
For deployers, the prohibition on AI-generated non-consensual intimate material and CSAM applies only where the deployer uses such AI systems for the purpose of generating or manipulating prohibited m Source
-
Special category data may be used for bias detection and correction by high-risk AI system providers only where the bias detection cannot be achieved using alternative data, subject to technical limit Source
-
The AI Office now has exclusive competence for supervision of AI systems based on general-purpose AI models where the model and system are developed by the same provider or by providers within the sam Source
-
The AI Office has exclusive competence for supervision of AI systems that constitute or are integrated into a very large online platform or very large online search engine within the meaning of the Di Source
-
Initial providers of high-risk AI systems must provide downstream providers with technical documentation, information about known limitations and failure modes, and targeted technical access for testi Source
-
Breaches of Article 25(2) or (4) now carry fines of up to 3% of worldwide annual turnover or €15 million, whichever is higher. Source
-
Rather than requiring providers and deployers to ensure AI literacy, the amended AI Act requires them to take measures to support the development of AI literacy among their staff and others dealing wi Source
-
The Digital Omnibus amendments are expected to be formally adopted with final approval anticipated in June 2026 and publication expected in July 2026. Source
-
Once adopted, the Digital Omnibus amendments will enter into force on the third day following publication in the Official Journal and will be binding in their entirety and directly applicable in all M Source