EU Withdraws AI Liability Directive Amid Global Regulatory Debates

AI Regulations: EU Withdraws AI Liability Proposal; US Seeks Comments on AI Action Plan | PYMNTS.com

The EU is withdrawing proposed regulations for handling victim's claims of harm caused by AI -- but it faces internal opposition. This week, the AI Liability Directive was taken off the 2025 work program of the European Commission, according to Euro News. The commission, which is the European Union's executive arm, cited lack of progress in negotiations on the directive. But legislators in the European Parliament's Internal Market and Consumer Protection Committee voted to keep working on the rules for artificial intelligence (AI) products. However, the leader of this issue, the Parliament's Legal Affairs committee, has not yet decided what to do. The AI Liability Directive was proposed in 2022 alongside the EU AI Act. While the AI Act served to assess risks posed by AI systems and penalize offending organizations, the AI Liability Directive would protect victims by making it easier to prove they were harmed, among other stipulations. The commission's change of heart was blasted by Axel Voss, a German member of the European Parliament. "Why the sudden U-turn? The answer likely lies in pressure from industry lobbyists who view any liability rules as an existential threat to their business models," Voss told the International Association of Privacy Professionals (IAPP). "Big Tech firms are terrified of a legal landscape where they could be held accountable for the harms their AI systems cause. Instead of standing up to them, the Commission has caved, throwing European businesses and consumers under the bus in the process," Voss said. The White House Office of Science and Technology Policy (OSTP) is asking the public to weigh in on policies for the "Artificial Intelligence (AI) Action Plan," being developed by the Trump administration. The plan seeks comments on which policies to prioritize to "enhance America's AI dominance and ensure that unnecessarily burdensome requirements do not hamper private sector innovation." The Request for Information notice, which was issued by the National Science Foundation on behalf of the OSTP, is seeking comments from the public until March 15, 2025. Stakeholders include academia; industry groups; state, local and tribal governments; private sector organizations and other interested parties. The request for comments is broad in scope, and comments can address any relevant AI policy topic. "The topics suggested by the RFI are wide-ranging and could impact nearly every facet of AI development and deployment," wrote Duane Pozza, Kathleen Scott and Lauren Lerman from the Wiley law firm in a blog post. They said areas of AI that could be affected include: AI is being increasingly used to help write books, create art, compose music and make movies. Thus, a common question that has been asked is whether creations that involved AI use are copyrightable. In a new report, the U.S. Copyright Office clarified that if AI is used to assist, rather than replace, human creativity, it would not affect its copyrightability. However, if a piece of creative work was entirely generated by AI, or where there isn't enough human control over the result, it is not copyrightable. A human using prompts alone to create the work is not enough. The U.S. Copyright Office also said assessing whether human effort is sufficient to constitute authorship will be analyzed on a case-by-case basis.

EU Withdraws AI Liability Directive Amid Regulatory Pushback

Disclaimer: This content generated by AI & may have errors or hallucinations. Edit before use. Read our Terms of use The European Commission withdrew its proposed AI Liability Directive (proposal aimed to hold AI developers accountable for damages caused by their systems) after facing criticism for excessive regulation and content moderation of AI at the AI Action Summit in Paris, France, Euractiv reported. The Commission explained the withdrawal in its final 2025 work programme, stating that "no foreseeable agreement" exists on the law. It will now decide whether to introduce a new proposal or adopt a different approach. The decision followed U.S. Vice-President JD Vance's strong opposition to stringent regulations on the rapidly evolving generative AI sector at the global summit. "We believe that excessive regulation of the AI sector could kill a transformative industry," Vance, 40, said. "He went as far as to label content moderation and regulation of generative AI as "authoritarian censorship," emphasising that the United States will continue to strive to be the leading AI stakeholder globally." The global summit in Paris aimed to highlight a more human-centric approach to AI governance. However, the EU and France's announcements of large-scale investments in AI development overshadowed this objective. The withdrawal of the AI Liability Directive appears to be a strategic move to position Europe as an attractive hub for innovation and global capital investment. It can also be seen as an attempt to appease the United States, which remains resistant to additional regulations that could hinder the emerging AI sector. French President Emmanuel Macron trod carefully, striking a balance between reducing bureaucratic hurdles for AI and implementing regulations to foster public trust in the technology. Realistically, the withdrawal of the AI Liability Directive appears pragmatic, as the proposal had been losing support in Europe over the past year following the adoption of the EU's AI Act -- the world's first comprehensive AI regulatory framework, passed in 2023. Additionally, Vance criticized the EU's Digital Services Act -- intended to create a safer digital environment -- and its stringent online privacy regulations, which he argued impose continuous legal costs on small businesses. In light of this, withdrawing the AI Liability Directive now seems like a prudent decision. Also Read: 1) European Union to introduce new Copyright rules for generative AI tools in its AI Act

EU withdraws its AI Liability Directive - What does this mean?

The Commission was unable to reach a 'foreseeable' agreement and plans to assess its future approach. After more than two years on the table, the European Commission has withdrawn the Artificial Intelligence (AI) Liability Directive after members failed to reach an agreement. The Directive, which consisted of non-contractual civil liability rules on AI, aimed to provide a clear and comprehensive structure for affected persons to claim compensation in the event they are harmed by the emerging technology. For example, this Directive would have simplified pursing legal action for those discriminated against by AI during a job recruitment process. The proposed rules built on existing ones for manufacturers to follow around unsafe product offerings, including a number of different strategies to modernise and adapt liability rules specifically for digital products. However, in its latest Work Programme published on 11 February, the Commission announced that it is withdrawing the Directive. It cited that "no foreseeable agreement" was reached and said that the Commission will be assessing its next approach. Although, the withdrawal wouldn't change much, said Forrester principal analyst Enza Iannopollo. "Many are either celebrating or mourning about the decision of the EU to withdraw the proposal for a new AI Liability Directive. Both are out of place. The AI Liability Directive was only one piece of the EU AI regulatory framework, which relies on a variety of other pieces of regulation, such as the GDPR and the EU AI Act. The latter is the core of the EU AI regulatory framework and organisations must stay committed to meet its requirements. "Some of the EU AI Act requirements are already enforceable and more will be in the next few months." The AI Liability Directive was initially tabled when the rules around product liability for the region were nearly 40 years old. However, a refreshed Product Liability Directive came into force in late December, which covers the latest technologies, including AI. Moreover, the EU's AI Act has been slowly coming into effect over the past few months after entering into force last August. Just days earlier, the Commission published non-binding guidelines on prohibited practices, which include banning companies from using an AI system to infer their employees' emotions, or assess a person's "risk" of committing a criminal offence. However, the Centre for Democracy and Technology, a non-profit civil rights organisation released a statement yesterday (12 February) expressing that the Directive's withdrawal represented a "significant setback" to the protection of victims against AI-induced harm. "Harms caused by AI systems and models are notoriously difficult to prove, owing to their complexity and lack of transparency. This leaves individuals with limited avenues to seek redress when they suffer harms induced by AI", said Laura Lazaro Cabrera, CDT Europe's Counsel and Director of the Equity and Data Programme. "[The Directive's] withdrawal is a departure from European values of transparency and accountability as well as fundamental rights, sending an alarming message that even the most basic procedural safeguards are fair play in the rush to embrace innovation" The EU's withdrawal came on the same day US vice-president JD Vance criticised the region's regulations around technology at the Paris AI Summit. He said that restrictions on AI would mean "paralysing one of the most promising technologies we have seen in generations". "The Trump administration is troubled by reports that some foreign governments are considering tightening screws on US tech companies with international footprints. America cannot and will not accept that, and we think it's a terrible mistake," Vance added. The Summit mobilised more than €300bn to ramp up AI investment in Europe in a bid to match the US's heavy-handed funding and increasingly loose regulations around AI. Don't miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic's digest of need-to-know sci-tech news.