{"id":146628,"date":"2026-08-12T09:19:05","date_gmt":"2026-08-12T09:19:05","guid":{"rendered":"https:\/\/my.legal500.com\/guides\/?post_type=comparative_guide&#038;p=146628"},"modified":"2026-08-12T09:44:23","modified_gmt":"2026-08-12T09:44:23","slug":"eu-artificial-intelligence","status":"publish","type":"comparative_guide","link":"https:\/\/my.legal500.com\/guides\/chapter\/eu-artificial-intelligence\/","title":{"rendered":"EU: Artificial Intelligence"},"content":{"rendered":"","protected":false},"template":"","class_list":["post-146628","comparative_guide","type-comparative_guide","status-publish","hentry","guides-artificial-intelligence","jurisdictions-eu"],"acf":[],"appp":{"post_list":{"below_title":"<div class=\"guide-author-details\"><span class=\"guide-author\">MPR Partners<\/span><span class=\"guide-author-logo\"><img src=\"https:\/\/my.legal500.com\/guides\/wp-content\/uploads\/sites\/1\/2019\/08\/mpr.jpg\"\/><\/span><\/div>"},"post_detail":{"above_title":"<div class=\"guide-author-details\"><span class=\"guide-author\">MPR Partners<\/span><span class=\"guide-author-logo\"><img src=\"https:\/\/my.legal500.com\/guides\/wp-content\/uploads\/sites\/1\/2019\/08\/mpr.jpg\"\/><\/span><\/div>","below_title":"<span class=\"guide-intro\">This country specific Q&amp;A provides an overview of Artificial Intelligence laws and regulations applicable in EU<\/span><div class=\"guide-content\"><div class=\"filter\">\r\n\r\n\t\t\t\t<input type=\"text\" placeholder=\"Search questions and answers...\" class=\"filter-container__search-field\">\r\n\t\t\t<\/div>\r\n\r\n\t\t\t\r\n\r\n\r\n\t\t\t<ol class=\"custom-counter\">\r\n\r\n\t\t\t\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">What is the legal definition of \u201cartificial intelligence\u201d in your jurisdiction, if any? If no definition exists, how do regulators or courts typically describe artificial intelligence?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The European Union\u2019s AI Act does not contain a general definition of artificial intelligence. It sets out instead the definitions of three AI-based technologies, namely AI systems, general-purpose AI systems and general-purpose AI models. It would therefore seem that AI systems are tantamount to AI in the EU terminology.<\/p>\n<p>This can be seen for example in a 2018 Communication on Artificial Intelligence for Europe, where the European Commission had noted that the term AI <em>\u201crefers to systems that display intelligent behaviour by analysing their environment and taking actions \u2013 with some degree of autonomy \u2013 to achieve specific goals<\/em>.\u201d<\/p>\n<p>The following year, the Independent High-Level Expert Group on Artificial Intelligence set up by the European Commission (the \u201c<strong>AI HLEG<\/strong>\u201d) defined AI systems as \u201c<em>software (and possibly also hardware) systems designed by humans that, given a complex goal, act in the physical or digital dimension by perceiving their environment through data acquisition, interpreting the collected structured or unstructured data, reasoning on the knowledge, or processing the information, derived from this data and deciding the best action(s) to take to achieve the given goal. AI systems can either use symbolic rules or learn a numeric model, and they can also adapt their behaviour by analysing how the environment is affected by their previous actions. As a scientific discipline, AI includes several approaches and techniques, such as machine learning (of which deep learning and reinforcement learning are specific examples), machine reasoning (which includes planning, scheduling, knowledge representation and reasoning, search, and optimization) and robotics (which includes control, perception, sensors and actuators, as well as the integration of all other techniques into cyber-physical systems).<\/em>\u201d<\/p>\n<p>The definition of AI systems has since evolved into that set out in the AI Act, as detailed below.<a href=\"#_ftnref1\" name=\"_ftn1\"><\/a><\/p>\n<p><strong>1.1. AI systems<\/strong><\/p>\n<p>According to the AI Act, an AI system is a \u201cmachine-based system that is designed to operate with varying levels of autonomy and that may exhibit adaptiveness after deployment, and that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments\u201d.<\/p>\n<p>This definition closely mirrors the revised definition adopted by the OECD in 2023 , as the definition initially proposed by the European Commission had faced criticism for deviating from the OECD\u2019s more technology-neutral one. This option however has been equally criticised on account that the OECD definition was not meant for legal use but as a public policy statement.<br \/>\nThe AI Act further defines the general-purpose AI (\u201cGPAI\u201d) system as an \u201cAI system which is based on a general-purpose AI model and which has the capability to serve a variety of purposes, both for direct use as well as for integration in other AI systems\u201d.<\/p>\n<p><strong>1.2. AI models<\/strong><\/p>\n<p>The AI Act defines a GPAI model as an \u201cAI model, including where such an AI model is trained with a large amount of data using self-supervision at scale, that displays significant generality and is capable of competently performing a wide range of distinct tasks regardless of the way the model is placed on the market and that can be integrated into a variety of downstream systems or applications, except AI models that are used for research, development or prototyping activities before they are placed on the market\u201d.<\/p>\n<p>The AI Act does not define AI models themselves. Such a definition has however been attempted by the OECD , according to which \u201dan AI model is a core component of an AI system used to make inferences from inputs to produce outputs\u201d.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">Has your country developed a national strategy for artificial intelligence? If yes, what progress has been made in its implementation? Are there plans for updates or revisions?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The EU has developed several AI strategies, including the Coordinated Plan on Artificial Intelligence , the AI Continent Action Plan and the Apply AI Strategy as well as sectoral strategies such as the European Strategy for Artificial Intelligence in Science &#8211; Paving the way for the Resource for AI Science in Europe (RAISE) and the Action Plan on Cybersecurity and Artificial Intelligence.<\/p>\n<p>Progress in the implementation of these strategies highly varies among Member States. However, as far as the European Commission is concerned, major milestones set in the Apply AI Strategy have been achieved by April 2026, with 19 AI factories being deployed, 13 AI Factory antennas providing regional access, the launch of an EU-India legal gateway office, and further progress on the AI Skills Academy.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">Has your jurisdiction adopted any AI-specific laws, regulations, voluntary standards, or ethical guidelines? If so, please provide a brief overview. If not, which existing laws could be\/are applied to artificial intelligence and the use of artificial intelligence, what are the main interpretive challenges, and are there any pending artificial intelligence legislative initiatives?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The AI Act was published in the Official Journal of the European Union on 12 July 2024 and entered into force on 1 August 2024. From this date<\/p>\n<p>The AI Act sets forth harmonised rules on the placing on the market, putting into service and use of AI systems within the European Union, whilst prohibiting certain AI practices and instating requirements for high-risk AI systems as well as transparency obligations for certain AI systems.<\/p>\n<p>The AI Act also contains rules on market monitoring, market surveillance, governance and enforcement, as well as measures supporting innovation, with a focus on SMEs.<\/p>\n<p>The AI Act was designed to become applicable gradually, at the following dates:<\/p>\n<ul>\n<li>2 February 2025 for general provisions, provisions on prohibited AI practices and on AI literacy;<\/li>\n<li>2 August 2025, for provisions on notifying authorities and notified bodies, general-purpose AI models, governance, confidentiality and penalties (except fines for providers of general-purpose AI models);<\/li>\n<li>2 August 2026 for all remaining provisions, except those related to the classification rules for high-risk AI systems (Article 6 paragraph (1) of AI Act) and corresponding obligations;<\/li>\n<li>2 August 2027, for classification rules and obligations pertaining to high-risk AI systems i.e., where AI systems are intended to be used as a safety component of a product or where the AI systems themselves are products covered by EU harmonised legislation, and the products in question are required to undergo third-party conformity assessments.<\/li>\n<\/ul>\n<p>As of 27 July 2026, the AI Omnibus (i.e., the EU regulation aimed at simplifying the AI Act) brought the following main amendments:<\/p>\n<ul>\n<li>the providers of AI systems, including general-purpose AI systems, which generate synthetic audio, image, video or text content and which have been placed on the market before 2 August 2026 are to comply with the transparency obligations set out at Article 50(2) of the AI Act (i.e. watermarking) by 2 December 2026;<\/li>\n<li>the rules related to the classification of AI systems as high-risk, the requirements for high-risk AI systems and the obligations of providers and deployers of high-risk AI systems and other parties will become applicable from 2 December 2027 as regards AI systems listed in areas such as biometrics, critical infrastructure, education and vocational training, employment, workers\u2019 management and access to self-employment, access to and enjoyment of essential private services and essential public services and benefits, law enforcement (in so far as their use is permitted under relevant Union or national law), migration, asylum and border control management (in so far as their use is permitted under relevant Union or national law, administration of justice and democratic processes), and from 2 August 2028 as regards AI systems embedded as safety components and covered by EU sectoral legislation on safety and market surveillance;<\/li>\n<li>a prohibition for AI systems to generate child sexual abuse material or to create images, videos or audio depicting an identifiable person\u2019s intimate parts or sexually explicit activities without their consent will apply from 2 December 2026.<\/li>\n<\/ul>\n<p>Pending the full application of the AI Act, the EU has also instituted the AI Pact, to help stakeholders prepare for the implementation of the AI Act, whereby organisations are encouraged to share best practices and to voluntarily implement the AI Act ahead of its full application.<\/p>\n<p>Furthermore, the European Commission has published the following guidelines and instruments to support the implementation of the AI Act:<\/p>\n<ul>\n<li>Guidelines on prohibited AI practices under the AI Act;<\/li>\n<li>Guidelines on the AI system definition of the AI Act;<\/li>\n<li>Guidelines on the scope of the obligations for the providers of GPAI models;<\/li>\n<li>GPAI Code of Practice;<\/li>\n<li>Template for the public summary of training content of GPAI models;<\/li>\n<li>Code of Practice on Transparency of AI-Generated Content;<\/li>\n<li>Guidelines on the implementation of the transparency obligations for certain AI systems under Article 50 of the AI Act.<\/li>\n<\/ul>\n<p>In May 2026, the European Commission submitted to public consultation the draft guidelines on the classification of high-risk AI systems (closing on 23 July 2026).<\/p>\n<p>Despite these guidelines, various provisions of the AI Act remain subject to interpretation, in particular as regards the definition of AI systems and high-risk AI systems.<a href=\"#_ftnref1\" name=\"_ftn1\"><\/a><\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">Are there legal requirements for artificial intelligence transparency, explainability, or audits? Are there obligations to disclose the use of artificial intelligence to customers\/clients?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p><strong>4.1. Transparency<\/strong><\/p>\n<p>The AI Act includes requirements on AI transparency. These are aimed at ensuring the development and usage of AI systems in a way that (i) allows proper traceability and explainability, (ii) makes humans aware that they communicate\/interact with an AI system, (iii) informs deployers of the capabilities and limitations of that AI system (supporting them in making an informed decision, interpreting the output and using it correctly) and (iv) informing affected persons about their rights concerning high-risk systems such as e.g. the right to an explanation referred to below, where applicable.<\/p>\n<p><strong>4.2. Explainability<\/strong><\/p>\n<p>The AI Act also provides for a right to an explanation in case of high-risk systems in areas such as biometrics, critical infrastructures, education and vocational training. The deployers of such AI systems must ensure this right where their decision is based mainly on the output of the AI system and that decision produces legal effects or it significantly affects a person in a way that they consider to have an adverse impact on their health, safety or fundamental rights. Where required as per the above, the explanations must be clear and meaningful as to enable affected persons to exercise their rights.<\/p>\n<p>The right to receive an explanation does not apply where (i) the EU law or EU law compliant national law provide exemptions from or restrict the obligation to provide explantions or (ii) a similar right exists in EU law, this being for example the case of high-risk systems processing personal data and providing decisions based solely on automated processing, where the right of access provided for under the General Data Protection Regulation prevails over the right to an explanation set out in the AI Act.<\/p>\n<p><strong>4.3. Audits<\/strong><\/p>\n<p>Finally, the AI Act enables the market surveillance authorities and notified bodies to carry out audits on AI systems. For example, the notified body will be able to carry out periodic audits of the quality management system put in place by the provider of AI systems. Auditing procedures are not regulated at EU level and should be adopted by way of national law in each EU Member State.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">Are there legal requirements or best practice expectations for human oversight and human-in-the-loop in artificial intelligence systems?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The AI Act requires that high-risk AI systems be designed\/developed in such a manner that natural persons can oversee their functioning. Thus, the provider of high-risk systems must implement proper human oversight measures before placing the system on the market\/putting the system into service. Such measures should include, amongst others:<\/p>\n<ul>\n<li>putting in place in-built operational constraints which cannot be overridden by the system and that respond to a human operator;<\/li>\n<li>ensuring that the person in charge of human oversight has the necessary competence, training and authority for this role;<\/li>\n<li>having mechanisms to guide and inform the person in charge of human oversight to make informed decisions to avoid negative consequences or risks, or stop the system if it does not perform as intended.<\/li>\n<\/ul>\n<p>The obligation to ensure human oversight aims to prevent and minimise risks to health, safety or fundamental rights that may emerge even when a high-risk AI system is used in accordance with its intended purpose or under conditions of reasonably foreseeable misuse.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">Are there specific legal or regulatory requirements addressing algorithmic bias, discrimination, or fairness in AI systems (including gender bias)?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The AI Act builds on the principles outlined in the 2019 Ethics guidelines for trustworthy AI developed by AI HLEG, which include diversity, non-discrimination and fairness. Thus, under the AI Act all AI systems must be developed and used in such a way as to include diverse actors and to promote equal access, gender equality and cultural diversity, whilst avoiding discriminatory impacts and unfair biases prohibited by EU law or national law.<\/p>\n<p>Further, data governance and management practices must be put in place to ensure that high-risk systems do not exhibit biases or discriminate against certain categories of people. These should include (i) performing an analysis with a view to identifying potential biases likely to either affect the health and safety of persons or to have a negative impact on fundamental rights or to cause discrimination prohibited under EU law and (ii) implementing appropriate measures aimed at detecting, preventing and mitigating possible biases identified following the performance of the above-mentioned analysis.<\/p>\n<p>For the purpose of (and only to the extent strictly necessary for) detecting biases and correcting the same, the AI Act allows providers of high-risk AI systems as a matter of substantial public interest, to process special categories of personal data, subject to appropriate safeguards for fundamental rights and freedoms of natural persons and only if certain criteria are cumulatively met, such as for example (i) the bias detection and correction cannot be effectively performed in the absence of special categories of data, (ii) technical limitations are implemented to prevent the re-use of the data, including state-of-the-art security and privacy-preserving measures, (iii) no other parties have access to the data, (iv) the data are deleted once the bias has been corrected or the retention period came to an end, whichever comes first.<\/p>\n<p>The AI Omnibus allows providers and deployers of all other AI systems and models and deployers of high-risk AI systems to process special categories of personal data when it is strictly necessary to ensure bias detection and correction. This will be applicable only if the possible biases are either likely to affect the health and safety of persons, have a negative impact on fundamental rights or lead to discrimination prohibited under EU law, subject to the same conditions and safeguards as mentioned above. Importantly, this possibility will not create an obligation for the providers and deployers of other AI systems and models and deployers of high-risk AI systems to conduct such bias detection and correction.<\/p>\n<p>As regards the obligation to ensure human oversight, deployers of high-risk AI systems must put in place measures enabling the person in charge of human oversight within their organization to be aware of the automation bias (i.e. the possible tendency of automatically relying or over-relying on the output produced by a high-risk AI system), especially where high-risk AI systems used to provide information or recommendations to support decisions taken by natural persons are concerned.<\/p>\n<p>Finally, information to be supplied by providers of GPAI models regardingdata used for training, testing and validation must include, where applicable, the methods employed in view of detecting identifiable biases.<a href=\"#_ftnref1\" name=\"_ftn1\"><\/a><\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">What legal frameworks apply to AI-related harm and defective artificial intelligence systems? Who can be held liable (developer, deployer, victim of the damage, others), how is liability allocated, and what burden of proof applies to victims?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The European Commission had intended to introduce a tailor-made AI liability directive, but the proposal in question was withdrawn in October 2025. In its place, Directive (EU) 2024\/2853 of the European Parliament and of the Council of 23 October 2024 on liability for defective products and repealing Council Directive 85\/374\/EEC (the \u201c<strong>New Product Liability Directive<\/strong>\u201d) &#8211; which entered into force on 8 December 2024 and must be transposed by EU Member States by 9 December 2026 &#8211; refers to AI systems at its recital (13), according to which AI systems fall under the concept of software regulated by the New Product Liability Directive.<\/p>\n<p>According to the said directive, manufacturers (which include providers of AI systems) bear the main liability for damages. Where AI providers are established outside the EU, the importers or authorised representatives thereof in the EU or so-called \u201cfulfilment service providers\u201d (where none of the former two exists) are liable instead. Where multiple economic operators may be held liable, their liability is joint and several.<\/p>\n<p>The claimant (i.e. the person having incurred harm) must prove the defectiveness of the product, the damage suffered (namely death or personal injury, including medically recognised damage to psychological health, damage to or destruction of any property, destruction or corruption of data that are not used for professional purposes) and the causal link between the defectiveness and that damage. Notwithstanding, the regime established by the New Product Liability Directive does not affect national law relating to compensation for damage under other liability regimes.<\/p>\n<p>The defectiveness of the AI system is presumed if:<\/p>\n<ul>\n<li>the defendant fails to disclose relevant evidence to counter the claim for compensation;<\/li>\n<li>the claimant demonstrates that the product does not comply with mandatory product safety requirements in the AI Act;<\/li>\n<li>the claimant proves that the damage was caused by an evident malfunction of the product which appeared during reasonably foreseeable use or under ordinary circumstances.<\/li>\n<\/ul>\n<p>According to the New Product Liability Directive, the defectiveness of the product and the causal link between the defectiveness and the damage are presumed where:<\/p>\n<ul>\n<li>it is excessively difficult for the claimant due to technical or scientific complexity to prove any of the above \u2013 which may well be the case of AI systems; and<\/li>\n<li>the claimant demonstrates that it is likely that the product is defective and that there is a causal link between defectiveness and the damage.<\/li>\n<\/ul>\n<p>Both presumptions referred to above can be rebutted by bringing evidence to the contrary.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">What cybersecurity obligations apply to AI systems?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The AI Act imposes a general obligation for high-risk AI systems to be <em>\u201ddesigned and developed in such a way that they achieve an appropriate level of accuracy, robustness, and cybersecurity, and that they perform consistently in those respects throughout their lifecycle\u201d<\/em>.<\/p>\n<p>It also provides that any technical solution intended to ensure the cybersecurity of high-risk AI systems must be appropriate to the relevant circumstances and risks.<\/p>\n<p>However, any high-risk AI system which is either certified or has a statement of conformity issued under the cybersecurity scheme set forth by Regulation 2019\/881 on the European Union Agency for Cybersecurity will be presumed to comply with the cybersecurity requirements set out in the AI Act if the cybersecurity certificate or statement of conformity covers such requirements.<\/p>\n<p>In addition, providers and deployers of high-risk AI systems have an obligation to report serious incidents to market surveillance authorities.<\/p>\n<p>Providers of GPAI models must equally ensure an adequate level of cybersecurity of the GPAI model with systemic risk and the physical infrastructure of such model.<\/p>\n<p>Furthermore, the NIS2 Directive will cover AI systems if they are part of the network and information systems of an essential\/important entity designated as such under the NIS2 Directive.<\/p>\n<p>Finally, the Cyber Resilience Act (which will enter into force on 11 December 2027) will also apply to AI systems where they amount to products with digital elements, such as for example. IoT devices. In this regard, the AI Omnibus expressly states that the AI Act and the Cyber Resilience Act complement each other. Thus, where high-risk AI systems fall within the scope of Cyber Resilience Act, such systems must also comply with the cybersecurity requirements set forth in the AI Act where applicable.<a href=\"#_ftnref2\" name=\"_ftn2\"><\/a><\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">Is the use of artificial intelligence insured and\/or insurable in your jurisdiction, including with cyber policies? Are there market trends, or limitations?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The use of artificial intelligence is insurable in the European Union. A number of insurers are offering insurance policies which such cover, for instance, as data poisoning, liability resulting from infringement of the AI Act, and usage rights infringement. Interest in acquiring AI insurance is high, as according to a Geneva Association report, 90% of businesses surveyed in 6 countries (including 2 EU countries, France and Germany) have recognised the need for AI-specific insurance, while over 66% are willing to pay higher premiums in order to cover AI risks.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">Can artificial intelligence be named as an inventor in a patent application filed in your jurisdiction? If not, what is the current legal position?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>According to the practice of the European Patent Office, only a natural person can be an inventor, so an artificial intelligence system cannot be named as an inventor in a European patent application. This perspective has crystallised on the occasion of the analysis of the DABUS cases, where, the European Patent Office has interpreted the relevant articles from the European Patent Convention and concluded that a non-human entity cannot hold rights or obligations, cannot be the original right holder to a patent, and therefore cannot transfer rights to the applicant.<\/p>\n<p>According to current EU legislative developments, there is no EU legislation in force that recognizes AI as inventor or allows it to be designated as such, and discussions on AI and intellectual property have so far left unchanged the requirement of a human inventor.<a href=\"#_ftnref1\" name=\"_ftn1\"><\/a><\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">Do images or works generated by and\/or with artificial intelligence benefit from copyright protection in your jurisdiction? If so, who is the authorship attributed to, and under what conditions?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The EU currently lacks legislation on the copyrightability of works generated by and\/or with artificial intelligence; accordingly, the answer to this question must be derived from the case law of the CJEU.<\/p>\n<p>According to well-established CJEU case law in cases such as <em>Infopaq International A\/S v. Danske Dagblades Forening<\/em> (C-5\/08) and <em>Eva-Maria Painer v. Standard VerlagsGmbH<\/em>, copyright requires the expression of an intellectual creation reflecting the author\u2019s personality, criteria that naturally cannot be satisfied without human input. Hence, images or works generated entirely autonomously by artificial intelligence (without meaningful human creative input) do not benefit from copyright protection.<\/p>\n<p>According to the European Parliamentary Research Service briefing \u201cCopyright of AI-generated works: Approaches in the EU and beyond\u201d, AI-assisted works can obtain copyright protection where there is significant human input and the final work is the result of the author\u2019s intellectual creation, even where AI tools were used in the process. Accordingly, AI-assisted works and images may qualify for copyright protection where they embody sufficient human creative input. This may be the case where a person meaningfully guides, selects, modifies or arranges the AI-generated output in a manner that reflects their own intellectual creation. In such circumstances, authorship would in any case be attributed to the natural person(s) whose creative choices are expressed in the final work and not to the AI.<\/p>\n<p>However, in recent national case law on copyright it has been decided that prompting alone does not normally confer copyright. In 2024, the Prague Municipal Court held that an AI-generated image could not benefit from copyright protection because it lacked the necessary element of human creative authorship. The court emphasised that only a natural person can be regarded as an author and that the mere act of submitting a prompt to an AI system does not automatically confer copyright in the generated output. The decision reinforces the broader European principle that copyright protection remains dependent upon sufficient human intellectual creation, even where AI tools are involved in the creative process.<a href=\"#_ftn2\" name=\"_ftnref2\"><\/a><\/p>\n<p>Although this decision is obviously not authoritative for other national courts, it is more likely than not that similar solutions will be adopted in other EU countries. This approach also appears to be reflected more broadly at the international level. In particular, the United States has reaffirmed the requirement of human authorship for copyright protection, with the courts confirming that works generated solely by AI are not eligible for copyright protection and that prompts, however sophisticated, are insufficient on their own to establish authorship.<\/p>\n<p>In light of the above, outputs that are generated entirely by AI, without human creative contribution, will fall outside the scope of copyright protection under the current EU legal framework. Conversely, where AI is used merely as a tool and the final output reflects sufficient human creative choices, such as directing, selecting, editing or arranging the AI-generated content, copyright protection may arise. However, human contribution may have to go well above mere prompting in order for copyright to be recognised and in any event, authorship will be attributed in such cases to the relevant natural person(s) whose creative contribution is expressed in the final work.<a href=\"#_ftnref1\" name=\"_ftn1\"><\/a><\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">What are the main issues to consider when using artificial intelligence systems in the workplace? Have any new regulations, or guidelines, been introduced regarding AI-driven hiring, performance assessment, or employee monitoring?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The main legal concerns surrounding the use of AI systems in the workplace relate to data protection, discrimination risks, transparency obligations and the limits imposed on automated decision-making. In practice, AI tools used for recruitment, employee evaluation or workplace monitoring will often involve profiling activities and, in some cases, decisions capable of producing significant effects on individuals, such as the rejection of job applications, refusal of promotion opportunities or the initiation of disciplinary measures.<\/p>\n<p>The AI Act classifies certain AI systems used in employment as \u201c<em>high-risk<\/em>\u201d. This includes AI systems intended to be used for recruitment, candidate selection, performance evaluation, promotion decisions, termination of employment or the monitoring of employee behaviour and performance.<\/p>\n<p>As a consequence of this classification, providers of such systems will be subject to extensive obligations before and after placing the systems on the market. These include conformity assessments, risk-management procedures, requirements relating to the quality and representativeness of training data, logging obligations, transparency measures and mechanisms ensuring meaningful human oversight. Employers using such systems will also have specific responsibilities, including the obligation to use the systems in accordance with the provider\u2019s instructions, monitor their operation and take corrective measures where discriminatory or otherwise problematic outcomes are identified.<\/p>\n<p>In addition, Article 22 GDPR prohibits decisions based solely on automated processing, including profiling, where they produce legal effects or similarly significant effects on a person, unless one of the exceptions under Article 22, paragraph (2) applies. In particular, the exception for contractual necessity is interpreted restrictively and applies only where the automated decision is objectively necessary for entering into or performing the contract, rather than merely convenient or efficient. Moreover, reliance on the data subject\u2019s explicit consent may be problematic in certain contexts, such as employment relationships, where the imbalance of power between the parties may prevent consent from being regarded as freely given.\u00a0 Even where one of these exceptions applies, the GDPR requires the implementation of safeguards, including the right to obtain human intervention, to express one\u2019s point of view and to challenge the decision.<\/p>\n<p>At the same time, the use of AI in employment relationships must comply with the Charter of Fundamental Rights of the European Union, including the right to respect for private life, the right to the protection of personal data and the principle of non-discrimination.<\/p>\n<p>In conclusion, the current EU framework does not prohibit the deployment of AI systems in employment-related contexts. Instead, it subjects such systems to a comprehensive set of safeguards designed to reconcile technological innovation with the protection of workers\u2019 fundamental rights. The overarching objective is to ensure that employment decisions remain fair, transparent and subject to effective human control, particularly where AI systems are used to influence or determine outcomes with significant consequences for individuals.<a href=\"#_ftnref1\" name=\"_ftn1\"><\/a><a href=\"#_ftnref2\" name=\"_ftn2\"><\/a><\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">What are the main privacy\/data protection issues arising from artificial intelligence development and use (including training data)? Have data protection authorities issued guidelines or rulings on artificial intelligence, and what are the key takeaways?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The main privacy and data protection concerns arising from the development and use of artificial intelligence in the EU stem from the fact that AI systems\/models are often developed, trained and deployed using very large amounts of personal data. As a result, the GDPR may apply at several stages of the AI system\/model\u2019s lifecycle, including during data collection, model training, testing and deployment.<\/p>\n<p>Controllers must identify a valid legal basis for processing personal data used to train AI systems\/models and must comply with the general principles set out at Article 5, paragraph (1) GDPR, including purpose limitation, data minimisation and transparency. This is particularly relevant since AI models are frequently trained on large and diverse datasets collected from multiple sources, including publicly accessible online content. However, personal data collected for one purpose cannot automatically be reused for unrelated AI training purposes without ensuring compliance with GDPR requirements.<\/p>\n<p>AI systems are also commonly used for profiling and automated decision-making, making Article 22 GDPR especially important. Article 22 GDPR grants individuals the right not to be subject to a decision based solely on automated processing, including profiling, where that decision produces legal effects concerning them or similarly significantly affects them, unless one of the exceptions set out in Article 22(2) applies.<\/p>\n<p>Another important issue concerns transparency. Many AI systems, particularly complex machine-learning models, operate in ways that are difficult for individuals to understand. Nevertheless, controllers remain subject to the information obligations set out at Articles 13 and 14 GDPR and must provide meaningful information about the logic involved, as well as the significance and envisaged consequences of the processing, where required under the GDPR. They must also implement mechanisms capable of facilitating the exercise of data subject rights, including access, rectification, erasure and objection, even where personal data has been incorporated into AI training datasets.<\/p>\n<p>The AI Act complements these GDPR obligations by introducing specific requirements for high-risk AI systems. According to Article 10 of the AI Act, training, validation and test datasets must be relevant, sufficiently representative and, as far as possible, free of errors and complete in view of the intended purpose of the system. Providers must also implement appropriate data governance and management practices designed to identify and mitigate potential bias and discriminatory outcomes. The AI Act recognises that, in narrowly defined circumstances, providers, and (once the AI Omnibus becomes applicable) deployers of AI systems may process special categories of personal data (see section 6 for information regarding such categories) for the purpose of detecting and correcting bias, subject to strict safeguards and compliance with the conditions set out at Article 9 GDPR.<\/p>\n<p>In December 2024, the European Data Protection Board (\u201c<strong>EDPB<\/strong>\u201d) issued an opinion concerning the processing of personal data in the context of AI models. Therein the EDPB clarified that the development and training of AI systems constitute processing activities fully subject to GDPR requirements and that organisations must ensure a lawful basis for processing at each stage of the AI lifecycle. Although not legally binding, EDPB opinions are highly persuasive and play an important role in ensuring the consistent interpretation and application of the GDPR across the EU.<\/p>\n<p>In addition, the EDPB\u2019s work on generative AI, including the ChatGPT taskforce report published in 2024 (the \u201c<strong>Report<\/strong>\u201d), set out its preliminary views on the interplay between the GDPR and AI. The Report emphasises in particular that the large-scale collection and use of personal data for the training of AI models must comply with the GDPR, including the requirements relating to lawful processing, transparency and the effective exercise of data subject rights.<\/p>\n<p>Data protection authorities have also consistently underlined that AI systems do not operate outside the existing legal framework. In practice, organisations developing or deploying AI are increasingly expected to conduct data protection impact assessments for high-risk processing activities, integrate data protection by design and by default into AI systems and adopt governance measures capable of reducing discriminatory or otherwise unlawful outcomes.<\/p>\n<p>Overall, the current EU approach is clearly that the use of AI does not diminish the application of data protection law. On the contrary, regulators increasingly expect organisations to address privacy, fairness, transparency and accountability considerations from the earliest stages of AI development and throughout the entire deployment process.<a href=\"#_ftnref1\" name=\"_ftn1\"><\/a><\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">How is data scraping regulated in your jurisdiction from an IP, privacy and competition perspective? Are there recent precedents addressing the legality of data scraping for training of artificial intelligence systems?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The legality of data scraping has been the subject of extensive regulatory and judicial scrutiny across the European Union, particularly in the context of artificial intelligence training. Since data scraping may engage different legal regimes, including intellectual property, data protection and competition law, the relevant legal framework and recent precedents are addressed separately below from each of these perspectives.<\/p>\n<ol>\n<li><strong>IP <\/strong><\/li>\n<\/ol>\n<p>From an EU IP perspective, data scraping is mainly assessed under copyright and database rights. Where the material being scraped is protected by copyright or by the sui generis database right, extraction or reuse can amount to infringement. The legality of data scraping for AI training in the EU is closely linked to the text and data mining (\u201c<strong>TDM<\/strong>\u201d) exceptions introduced by the DSM Copyright Directive. TDM is defined as any automated analytical technique aimed at analysing text and data in digital form in order to generate information such as patterns, trends and correlations. However, data scraping and text and TDM are distinct concepts, to the extent that aata scraping refers to the automated collection of content. Scraping is often a preliminary step in TDM and AI training. Under Articles 3 and 4 of the DSM Copyright Directive, certain acts of reproduction and extraction carried out for TDM purposes may benefit from copyright exceptions. In the context of AI training, Article 4 of the DSM Directive has become the principal legal basis relied upon by many providers of generative AI systems, subject to the right of rights-holders to opt out. This approach is reinforced by Article 53, paragraph (1), (c) of the AI Act, which requires providers of general-purpose AI models to identify and comply with rights reservations expressed under Article 4, paragraph (3) DSM Directive. Nevertheless, the extent to which generative AI training falls within the notion of TDM remains the subject of considerable academic, judicial and regulatory debate, and several Member States have expressed the view that certain AI training activities may go beyond the scope of the TDM exceptions.<\/p>\n<p>The most significant development at EU level is the pending case of Like Company v Google Ireland (Case C-250\/25), which constitutes the first reference to the Court of Justice of the European Union concerning the application of EU copyright law to generative artificial intelligence systems. The dispute arose after a Hungarian press publisher alleged that Google&#8217;s AI chatbot generated content closely resembling one of its press articles.<\/p>\n<p>The referring Hungarian court has asked the CJEU to clarify several fundamental issues regarding the interaction between generative AI and copyright law. In particular, the Court has been asked to determine whether the use of copyrighted content during the training of large language models constitutes an act of reproduction within the meaning of the InfoSoc Directive (Directive 2001\/29\/EC), whether AI-generated outputs that reproduce elements of protected works may amount to a communication to the public, and whether commercial AI training activities can benefit from the text and data mining exception provided by Article 4 of the DSM Directive (Directive (EU) 2019\/790).<\/p>\n<p>The case is expected to clarify whether the training and operation of generative AI systems on copyrighted content can take place without the authorisation of rightsholders, thereby shaping the future balance between innovation and copyright protection in the European Union. As the proceedings are still pending, no definitive guidance has yet been provided by the CJEU. Nevertheless, the forthcoming judgment is expected to become one of the most influential decisions in the field of AI and copyright in Europe, as it should provide an authoritative interpretation of the EU copyright framework and will be binding on courts across all Member States.<\/p>\n<ol start=\"2\">\n<li><strong>Data Protection <\/strong><\/li>\n<\/ol>\n<p>From a data protection perspective, data scraping falls under the obligations imposed by the GDPR. Hence, organisations using web scraping must ensure they have a lawful basis under the GDPR for processing personal data. When web scraping captures special categories of personal data, such as health information, additional conditions under Article 9 GDPR apply. These conditions include the necessity for explicit consent or meeting specific conditions, such as processing for substantial public interest or scientific research purposes. Additionally, the AI Act expressly prohibits the untargeted scraping of facial images from the internet or CCTV footage for the creation of facial recognition databases.<\/p>\n<p>A recent development in this area is the adoption by the EDPB on July 7, 2026, of Guidelines No 3\/2026 on web scraping in the context of generative AI, which are currently open for public consultation. The Guidelines set out the EDPB&#8217;s approach on the application of the GDPR to web scraping for AI training, addressing issues such as the identification of an appropriate legal basis, compliance with the principles of purpose limitation, transparency, data minimisation and accuracy, the processing of special categories of personal data, the allocation of roles between controllers and processors, and the safeguards expected throughout the scraping and AI training process.<\/p>\n<ol start=\"3\">\n<li><strong>Competition Law <\/strong><\/li>\n<\/ol>\n<p>From a competition law perspective, data scraping is not inherently unlawful under EU law. However, competition issues may arise depending on the market position of the parties, the nature of the data concerned, the conditions governing access to that data and the potential of the conduct to produce exclusionary or exploitative effects. In particular, competition law may become relevant where access to data is necessary for competing in a downstream or adjacent market. Large digital platforms may hold significant volumes of data which are difficult or impossible for competitors to replicate. In such circumstances, restrictions on access to data, discriminatory access conditions or technical measures preventing interoperability may be assessed under Article 102 TFEU if the relevant undertaking holds a dominant position and the conduct is capable of excluding competitors or strengthening market power.<\/p>\n<p>At the same time, competition law does not create a general right to scrape third-party websites or databases. A refusal to grant access to data will not automatically amount to an abuse of dominance. The assessment remains fact-specific and requires an analysis of dominance and various other conditions depending on the particular circumstances of the case.<\/p>\n<p>Data scraping may also be relevant in the context of the Digital Markets Act, which imposes specific obligations on designated gatekeepers. While the DMA does not create a general permission to scrape protected content, it seeks to address structural data-related advantages held by major digital platforms, including by imposing obligations relating to data access, data portability, interoperability and non-discriminatory treatment in certain circumstances. Therefore, in digital markets, the legality of restrictions on data access may need to be assessed not only under Articles 101 and 102 TFEU, but also under the DMA where the relevant undertaking is a designated gatekeeper and the conduct falls within the scope of its obligations.<\/p>\n<p>Conversely, data scraping may also amount to unfair competition under the national laws of the EU Member States, even where publicly accessible.<a href=\"#_ftnref1\" name=\"_ftn1\"><\/a><\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">To what extent is the prohibition of data scraping in the terms of use of a website enforceable?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The enforceability of contractual prohibitions on data scraping depends on the applicable national law and the specific circumstances of the case. Nevertheless, recent national case law suggests that such clauses may be effectively enforced in certain countries.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">Does your country have a regulator or authority responsible for supervising the use and development of artificial intelligence? What are its powers and enforcement tools?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>At the EU level, the European AI Office, established within the European Commission, is responsible for overseeing the implementation of the AI Act, particularly in relation to general-purpose AI models. It is empowered to monitor compliance, request information and documentation, conduct evaluations, investigate potential infringements, and enforce the AI Act, including through the imposition of administrative fines where appropriate.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">How widespread is the adoption of artificial intelligence in businesses in your jurisdiction, and which sectors are leading?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>AI adoption across the European Union has accelerated significantly in recent years, driven by rapid technological advances, increased investment, and the growing availability of generative AI tools. While levels of adoption still vary between Member States and sectors, AI is now widely recognised as a strategic priority for businesses operating in the EU. This trend has been reinforced by the EU\u2019s broader digital transformation agenda and the introduction of the EU AI Act, which seeks to establish a harmonised regulatory framework for trustworthy AI.<\/p>\n<p>According to recent Eurostat data, larger enterprises remain the leading adopters of AI technologies, although uptake among small and medium-sized enterprises (\u201cSMEs\u201d) is also steadily increasing. Businesses are primarily deploying AI for process automation, data analytics, customer interaction, cybersecurity, and operational efficiency. Generative AI applications, including large language models and AI-assisted productivity tools, have seen particularly rapid growth in recent years.<\/p>\n<p>The information and communication, financial services, healthcare, manufacturing, and retail sectors are currently among the leading adopters of AI within the EU.<\/p>\n<p>The technology sector continues to drive innovation and deployment, particularly in software development, cloud services, cybersecurity, and digital platforms. AI is extensively used for coding assistance, predictive analytics, fraud detection, and customer support automation.<\/p>\n<p>Financial institutions across the EU have also adopted AI at scale, especially in banking, insurance, and fintech. Common use cases include algorithmic risk assessment, anti-money laundering (\u201cAML\u201d) monitoring, fraud detection, credit scoring, and automated customer service solutions. Given the highly regulated nature of the sector, financial services firms are also among the most active in developing AI governance and compliance frameworks.<\/p>\n<p>In healthcare and life sciences, AI adoption has expanded rapidly in diagnostics, medical imaging, drug discovery, and healthcare administration. Hospitals and pharmaceutical companies increasingly rely on machine learning tools to improve efficiency and support clinical decision-making, although such applications are subject to stringent requirements under EU data protection and medical device regulations.<\/p>\n<p>The manufacturing sector is another key driver of AI deployment, particularly in Germany, France, Italy, and other industrial economies. Manufacturers are integrating AI into predictive maintenance, robotics, supply chain optimisation, and quality control processes as part of broader Industry 5.0 initiatives.<\/p>\n<p>Retail and e-commerce businesses are increasingly utilising AI for personalised marketing, customer analytics, and demand forecasting. Generative AI tools are also being used to automate content creation and enhance customer engagement.<\/p>\n<p>Overall, AI adoption in the EU is becoming both broader and more sophisticated. While the technology sector and heavily data-driven industries currently lead the market, AI deployment is expanding across virtually all areas of the European economy as organisations seek efficiency gains, innovation opportunities, and competitive advantages.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">How is artificial intelligence used in the legal sector, by lawyers and\/or in-house counsels? Are AI-driven legal tools widely adopted, and what are the main regulatory concerns?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>Artificial intelligence is increasingly used in the legal sector across a broad range of activities by both law firms and in house legal departments. In practice, AI tools are commonly employed for document review, legal research, contract analysis, due diligence, litigation support, compliance monitoring, e discovery, translation, and the drafting or summarisation of legal documents, often on the basis of large datasets and advanced machine learning techniques . According to the Thomson Reuters Institute\u2019s 2025 Generative AI in Professional Services Report, 26% of professionals now use generative AI at work, up from 14% in 2024, which represents almost a twofold increase year on year. Usage rates were highest among law firm attorneys, followed by in-house counsel and then government lawyers.<\/p>\n<p>Across the EU, AI is generally viewed as an assistive tool rather than a substitute for professional legal judgment. Lawyers naturally remain fully responsible for the legal advice provided to clients and for ensuring compliance with EU data protection and professional secrecy obligations.<\/p>\n<p>Although GenAI tools can improve efficiency and support the delivery of legal services, their use may conflict with lawyers\u2019 professional obligations in several ways. Hallucinations can lead GenAI systems to generate fictitious case law, non existent judicial opinions or entirely invented legal arguments, while bias and \u201csycophancy\u201d may cause outputs to reflect or reinforce existing prejudices or simply tell users what they appear to want to hear, resulting in misleading or unbalanced content. A lack of transparency makes it difficult for lawyers to understand how outputs are produced and to verify their reliability. Further concerns include intellectual property issues around the use of copyrighted or unlicensed material in training data and the ownership or infringement risks related to inputs and outputs, as well as increased cybersecurity and fraud risks (for example, deepfakes, synthetic identities and AI driven scams) that can compromise sensitive information and damage reputations.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">What are the 5 key challenges and the 5 key opportunities raised by artificial intelligence for lawyers in your jurisdiction?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>Artificial intelligence raises both significant opportunities and important challenges for lawyers and legal practice.<\/p>\n<p><strong>The five key opportunities are:<\/strong><\/p>\n<p>(i) increased efficiency and productivity; AI tools can automate repetitive and time-consuming tasks such as document review, drafting, translation, document summarisation and legal research, allowing lawyers to work more quickly and efficiently;<\/p>\n<p>(ii) enhanced legal research and analysis; AI systems can rapidly analyse large volumes of legal materials, identify relevant case law, detect legal trends and assist with compliance checks, potentially improving the quality and speed of legal analysis;<\/p>\n<p>(iii) cost savings and better resource allocation; AI may reduce operational costs, accelerate case handling and allow lawyers to focus more on strategic and qualitative legal work rather than routine administrative tasks;<\/p>\n<p>(iv) improved client service; AI tools may improve responsiveness, facilitate better workload and cost estimation, and make legal services more accessible to clients by lowering costs and increasing availability;<\/p>\n<p>(v) development of new legal specialisations and working methods; AI is transforming legal practice by encouraging lawyers to develop technological competence and focus increasingly on higher-value advisory and strategic functions.<\/p>\n<p><strong>The five key challenges are:<\/strong><\/p>\n<p>(i) maintaining confidentiality and ensuring compliance with privacy and data protection obligations; lawyers must ensure that the use of AI systems does not result in the inadvertent disclosure, retention or misuse of confidential client information or personal data;<\/p>\n<p>(ii) ensuring the accuracy and reliability of AI-generated outputs; lawyers must address the risk that AI systems may produce inaccurate, misleading or entirely fabricated legal information, including fictitious case law or incorrect legal reasoning;<\/p>\n<p>(iii) preserving professional independence and objective legal judgment; lawyers must remain vigilant against the risk that AI systems may reproduce societal biases or excessively align with users\u2019 assumptions, potentially influencing legal analysis and decision-making;<\/p>\n<p>(iv) protecting legal practice against cybersecurity, intellectual property and fraud-related threats; lawyers must manage the increasing risks associated with data breaches, misuse of copyrighted material, deepfakes, impersonation, synthetic identity fraud and other AI-enabled attacks;<\/p>\n<p>(v) maintaining professional competence and adequate human oversight; lawyers must ensure that they understand the capabilities and limitations of AI tools and continue to exercise independent legal review and supervision over AI-generated work product.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\t\t\t\t\t<li class=\"question-block filter-container__element\">\r\n\t\t\t\t\t\t<h3 class=\"filter-container__match-html\">Where do you see the most significant legal developments in artificial intelligence in your jurisdiction in the next 12 months? Are there any ongoing initiatives that could reshape AI governance?<\/h3>\r\n\t\t\t\t\t\t<button id=\"show-me\">+<\/button>\r\n\t\t\t\t\t\t<div class=\"question_answer filter-container__match-html\" style=\"display:none;\"><p>The most significant legal changes around artificial intelligence over the next year are likely to centre on how the AI Act is put into practice and enforced.<br \/>\nParticular attention is expected to focus on high-risk AI systems, transparency requirements, general-purpose AI models, and the role of the European AI Office in overseeing enforcement and ensuring consistent application of the AI Act across the EU. In practice, one of the main developments will likely be the issuance of guidance, standards and compliance frameworks intended to help organisations understand how the AI Act should be applied in specific operational contexts.<\/p>\n<p>At the same time, many of the AI Act provisions will require further clarification through regulatory guidance, harmonised standards and, ultimately, judicial interpretation. For example, greater legal certainty will be needed regarding the scope of the definition of an AI system, including the distinction between AI systems and traditional software, as well as the classification of certain AI applications under the AI Act&#8217;s risk-based framework. Additional clarification is also expected in relation to the obligations applicable to providers and deployers of AI systems, particularly where responsibilities overlap across complex AI value chains.<\/p>\n<p>Beyond the AI Act itself, a number of legal issues are likely to continue evolving through case law and regulatory practice. These include the application of intellectual property rules to AI-generated content and AI training, the lawfulness of web scraping and text and data mining for model development, the ownership and protection of AI-generated outputs, and the interaction between the AI Act and existing legal frameworks, including the GDPR, copyright law, consumer protection legislation and the revised product liability regime, among other things. As disputes involving AI technologies become more frequent, judicial decisions are expected to play an increasingly important role in shaping the practical interpretation of these issues.<\/p>\n<p>Consequently, the next 12 months are likely to be defined not only by the transition from legislative adoption to operational compliance and enforcement, but also by the gradual development of a more coherent and predictable legal framework through guidance issued by the European Commission and the AI Office, the adoption of harmonised technical standards, and the emergence of national and EU-level case law addressing unresolved questions raised by the deployment of AI technologies.<\/p>\n<\/div>\r\n\r\n\r\n\t\t\t\t\t<\/li>\r\n\r\n\t\t\t\t\r\n<div class=\"word-count-hidden\" style=\"display:none;\">Estimated word count: <span class=\"word-count\">8577<\/span><\/div>\r\n\r\n\t\t\t<\/ol>\r\n\r\n<script type=\"text\/javascript\" src=\"\/wp-content\/themes\/twentyseventeen\/src\/jquery\/components\/filter-guides.js\" async><\/script><\/div>"}},"_links":{"self":[{"href":"https:\/\/my.legal500.com\/guides\/wp-json\/wp\/v2\/comparative_guide\/146628","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/my.legal500.com\/guides\/wp-json\/wp\/v2\/comparative_guide"}],"about":[{"href":"https:\/\/my.legal500.com\/guides\/wp-json\/wp\/v2\/types\/comparative_guide"}],"wp:attachment":[{"href":"https:\/\/my.legal500.com\/guides\/wp-json\/wp\/v2\/media?parent=146628"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}