EU AI Act Enforces New Rules for Safer, Transparent AI

New EU AI Rules Take Effect

As of yesterday, a significant phase of the European Union’s Artificial Intelligence (AI) Act has officially come into force, introducing mandatory safety and transparency standards for developers and providers of AI technologies. This move requires governments across the EU, including Ireland, to establish regulatory structures to oversee compliance among companies offering AI tools.

The timing of these new rules contrasts sharply with developments in the United States, where the government is taking steps to reduce regulatory burdens in a bid to dominate the global AI race.

Transparency Now a Must for AI Systems

At the digital agency All human in Dublin, developers are currently building an AI-powered energy price comparison tool. The system uses a chatbot that not only speaks with an Irish accent but also sounds convincingly human. Under the new EU regulations, such systems must clearly disclose to users when they are interacting with AI rather than a real person.

“With the sophistication of AI, artificial voices can be so nuanced that users may believe they’re speaking to a human,” said John Mitchell, CEO of All human. “Thanks to the AI Act, companies must now declare what is real and what is not.”

Ban on High-Risk AI and Mandatory Compliance

The AI Act, which was adopted last year, is being implemented in phases. The latest phase prohibits AI systems that pose a significant threat to public safety, individual rights, and livelihoods. It also lays out comprehensive requirements for transparency, accountability, and safety.

Maureen Daly, a partner at law firm Pinsent Masons Ireland specializing in intellectual property and AI, emphasized the urgency for businesses to start preparing. “This applies not only to tech giants like Google and Meta but also to startups and SMEs,” she explained. “Everyone with AI systems must comply.”

Daly also warned of significant financial penalties for non-compliance. “While the European Commission won’t begin issuing fines until August 2, 2026, that doesn’t mean companies should delay preparations,” she noted. “Fines can reach up to 3% of global annual turnover or €15 million.”

Government Responsibilities and Delays

EU Member States are now obligated to establish enforcement mechanisms and designate competent authorities. Ireland, like other member states, must legislate penalties for violations and support regulatory bodies with necessary resources.

Dr. Kris Shrishak, a senior fellow at the Irish Council for Civil Liberties (ICCL), criticized the Irish Government for not acting sooner. “This should have started much earlier,” he said. “We need urgency, not just in setting up regulatory bodies, but also in providing them with resources.”

Dr. Shrishak added that regulatory efforts should also focus on fostering responsible AI innovation. “Sandboxes should be created to help guide Irish companies. It’s not only about stopping harmful AI but also about encouraging beneficial applications.”

Responding to these concerns, a spokesperson for the Department of Enterprise stated that Ireland has met all its obligations to date. “A significant body of work is ongoing, including workshops and collaboration with other EU Member States,” the spokesperson said.

US vs EU: Diverging Approaches to AI Regulation

While the EU tightens its regulatory grip, the United States is moving in the opposite direction. President Donald Trump has pledged to eliminate bureaucratic obstacles to ensure that the U.S. wins the AI race.

This divergence has sparked debate over whether strict regulations could hinder innovation in Europe. Caroline Dunlea, Chairperson of Digital Business Ireland (DBI), acknowledged this concern. “Bridging the competitiveness gap between the EU and countries like the U.S. and China is vital,” she said. “We must strike a balance to avoid overregulation.”

John Mitchell of All human echoed this sentiment. “There needs to be guardrails, and I think the EU has struck a good balance between being progressive and protective,” he said. “In the long run, that’s the best approach for an industry that currently lacks regulation.”

A special radio segment discussing the new AI rules is scheduled to air on This Week on RTE Radio 1 at 1pm today.


This article is inspired by content from Original Source. It has been rephrased for originality. Images are credited to the original source.

Subscribe to our Newsletter