Historic EU AI Regulations Secure Backing from EU Nations

The European Union is poised to enact historic EU AI Regulations on artificial intelligence, marking a significant milestone in the digital age. This legislative endeavor aims to create a balanced ecosystem where AI can thrive responsibly. By establishing these regulations, the EU seeks to set a framework that ensures AI's advancements are united with ethical standards and human rights. This move could potentially set a global benchmark, influencing how AI is governed worldwide. As the EU charts a course through this new frontier, it endeavors to foster innovation while upholding the principles of transparency, accountability, and fairness in the rapidly evolving landscape of artificial intelligence.

EU Reaches Preliminary AI Legislation Agreement

On Friday, 17 May 2024, the European Union and European Parliament reached a preliminary agreement on landmark artificial intelligence laws (AI), including the use of AI for biometric surveillance by governments and the regulation of AI systems like ChatGPT, as the EU moves closer to becoming the world’s first major power to enactEU AI Regulations on AI.

Marathon Discussions Precede Agreement

The political agreement reached between EU member states and the European Parliament followed a nearly 15-hour meeting on Friday after a nearly 24-hour discussion the day before. The two sides are expected to negotiate further on the details of the EU AI Regulations in the days ahead, which could alter its form.

Europe Leads with Global Standards

Europe has established itself as a leader in the world, recognizing the importance of being a global standard-setter. "I believe this is a momentous occasion," said Thierry BRETON, European Commissioner, at a press conference to announce the agreement.

Transparency Requirements for AI Models

The agreement requires foundation models like ChatGPT, as well as GAI, to meet transparency requirements before they are released to the market. This includes technical documentation, compliance with EU AI Act copyright legislation, and detailed summaries of training content.

Assessments and Regulations for High-Impact AI

High-impact foundation models with a systemic risk will be required to conduct model assessments, assess and reduce systemic risks, perform peer-to-peer testing, notify the European Commission of serious incidents, guarantee cybersecurity, and report on energy efficiency.

Restrictions on Biometric Surveillance

Systemic risk-based pay-as-you-go (GAPI) applications may be subject to codes of practice. According to the AI Governance may only deploy real-time biometrics in public areas for:

• Victimisation of certain crimes

• Prevention of genuine, present, or foreseeable threats such as terrorist attacks

• Sight-searing of individuals suspected of serious crimes

Prohibitions and Consumer Rights

Cognitive behavioral manipulation (CBM) is prohibited, as well as the untargeted scraping of facial images from the internet or CCTV footage. Social scoring and biometrics systems for inferring political, religious, and philosophical beliefs, as well as sexual orientation and race, are prohibited. Consumers will have the right to file a complaint and receive meaningful explanations. Violations will be subject to fines ranging from EUR 7.5 million (1.5%) to EUR 35 million (7% of global turnover).

Industry Reactions to the New European Union AI Policy

Digital Europe, a business group, described the rules as "yet another burden for companies" on top of other new legislation. "We have an agreement, but at what price? We were fully in favor of a risk-based model based on the use of AI, rather than the technology itself, and the last-minute effort to regulate foundation models appears to have thrown this plan into disarray," said Digital Europe's Director General.

Explore here the 10 things to know about the EU AI Act.

Criticism from Privacy Advocates

Ella Jakubowska, a senior policy advisor at European Digital Rights, a privacy rights group, was equally critical. "It’s difficult to be enthusiastic about a law that has made live public facial recognition legally available across the EU for the first time," she said. "The Parliament did its best to limit the harm, but the overall package around biometrics and profiling is lukewarm at best."

Legislation Timeline and Global Context

The legislation is set to come into effect early next year once it has been formally ratified by both political parties, and should then apply two years later. Governments around the globe are trying to weigh the benefits of artificial intelligence (AI) that can conduct human conversations, answer questions, and write computer code against the need to impose controls.

AI Companies Continue to Innovate Amid New Rules

Europe’s ambitious AI rules come at a time when AI companies, including Microsoft (MSFT) investor OpenAI, continue to find new applications for their technology, drawing both praise and criticism. Google parent Alphabet (GOOGL.O.) on Thursday unveiled a new artificial intelligence model, Gemini, that competes with OpenAI.

Potential Global Impact of EU AI Legislation

The EU legislation could serve as a model for other governments and provide an alternative to the US’s light touch approach and China’s provisional rules.

Conclusion

As the EU endures revolutionary EU AI Regulations, it stands at the lead of a global paradigm shift. These regulations’ goal is to harness AI's potential while conserving citizens' rights, setting a precedent for responsible technology use. The world watches as Europe directs this uncharted territory, potentially influencing international standards. The success of these regulations could motivate similar frameworks globally, ensuring AI caters to humanity with justice and integrity. Amidst this, AI companies continue to invent and adapt to these changes. The EU's bold steps today could shape the moral landscape of AI for generations to come, demonstrating a commitment to progress without compromising on values.