AI Daily News: European CEOs Urge Delay of AI Act Implementation

European CEOs Urge Delay of AI Act Implementation

Top executives from 44 major European companies, including Airbus, BNP Paribas, Carrefour, and Philips, have called on the European Commission to pause the implementation of the EU's landmark Artificial Intelligence (AI) Act, scheduled to come into effect in August 2025. In an open letter to Commission President Ursula von der Leyen, the CEOs argue that the Act’s complex and overlapping regulations jeopardize Europe's competitiveness in the global AI arena, potentially stifling innovation and deterring investors. The criticism coincides with an ongoing reevaluation by EU and industry stakeholders to simplify the rules, especially the drafting of a 'code of practice' meant to guide firms in complying with the Act. Many provisions of the law are not yet in effect, but businesses fear the unclear regulations could burden smaller companies and hinder broader AI deployment. EU officials have committed to finalizing the code of practice by August and are considering regulatory simplification. Meanwhile, European AI start-ups and investors have also voiced strong concerns, describing the legislation as hasty and damaging. The European Commission maintains its commitment to ensuring AI safety and regulatory harmonization, while indicating that further changes to the digital framework remain possible. (ft.com)

Meta's Aggressive Investment in AI Talent

Mark Zuckerberg has launched Meta Superintelligence Labs, investing billions to recruit top AI talent. Compensation packages exceed $100 million annually, with up to $300 million over four years. This talent war is reshaping the AI landscape and complicating U.S. government efforts to attract top scientists, especially as Chinese companies gain global ground. Meta’s investment includes $14 billion in Scale AI and key hires from GitHub and OpenAI, reflecting Zuckerberg's belief in AI’s massive future revenues. (axios.com)

Microsoft's AI Outperforms Doctors in Diagnosing Diseases

Microsoft's AI-driven medical tool, the AI Diagnostic Orchestrator (MAI-DxO), dramatically outperforms human doctors in diagnosing complex medical cases. In a study analyzing 300 challenging cases from the New England Journal of Medicine, MAI-DxO correctly diagnosed 85% of cases, compared to just 20% by 21 general-practice doctors from the U.S. and U.K. The AI also achieved diagnoses at an average of 20% lower cost by efficiently ordering fewer, more relevant tests. Designed to replicate the iterative diagnostic process of clinicians, the system uses multiple advanced AI models and a virtual 'Orchestrator' to simulate expert consultations. While still in the research phase and not subject to clinical regulation, MAI-DxO could potentially reduce medical errors, lower costs, and improve healthcare accessibility globally. Experts stress, however, that more real-world clinical trials are needed. Comparatively, Google’s diagnostic AI recently achieved a 59% accuracy rate. Microsoft sees the future application of such AI systems not only as clinical aids but as tools to democratize expert healthcare access worldwide. (time.com)

U.S. Senate Removes AI Regulatory Ban from GOP Bill

The U.S. Senate overwhelmingly voted 99-1 to remove a proposed 10-year ban on state regulation of artificial intelligence (AI) from a major Republican bill that included tax breaks and spending cuts. The proposal, aimed at centralizing AI oversight at the federal level by withholding funding from states unless they refrained from regulating AI, faced bipartisan opposition from governors, attorneys general, and AI safety advocates. Efforts to revise the provision to a five-year moratorium with exemptions for child safety and artist protections failed after Sen. Marsha Blackburn (R-TN) and Sen. Maria Cantwell (D-WA) successfully introduced an amendment to strike it. Critics argued the moratorium would have given AI companies unchecked freedom and stripped states of the ability to address emerging harms. Parents of children harmed by AI tools and Republican governors, including Arkansas Gov. Sarah Huckabee Sanders, strongly opposed the ban, emphasizing states' roles in regulating technology. Despite support from some tech leaders and Sen. Ted Cruz (R-TX), who attempted a last-minute compromise, the Senate rejected the measure, preserving states' rights to regulate AI technologies. (apnews.com)

Amazon Enhances AI Capabilities with Covariant Acquisition

Amazon has hired the founders of AI robotics startup Covariant, Pieter Abbeel, Peter Chen, and Rocky Duan, along with approximately a quarter of the startup's employees. The move bolsters Amazon's AI capabilities and talent pool, providing a boost to its robotics initiatives. Additionally, Amazon has signed a non-exclusive license to use Covariant's robotic foundation models, which are based on the 'Covariant Brain' platform. Covariant, a Bay Area startup, develops AI for advanced warehouse robotics systems, enabling robots to see, reason, and act on their surroundings. Their technology automates various warehouse tasks, including order picking, item induction, and depalletization. The company has raised $222 million in funding to date and counts healthcare supply manufacturer McKesson and German retail giant Otto Group among its customers. (dailyaitimes.com)