The Download: L'IA cinese divide la Casa Bianca, e un record di pagamento del copyright

The U.S. government is facing internal divisions over how to respond to the rise of Chinese AI models, particularly following the release of Kimi K3 by Moonshot, a Chinese AI company. Several advisors to former President Donald Trump who focus on artificial intelligence have publicly criticized major U.S. AI firms such as Anthropic and OpenAI, accusing them of being overly cautious or politically aligned. David Sacks, a former Trump adviser, described Anthropic’s models as “lobotomized” and “woke,” while Emil Michael, a senior Pentagon official, referred to OpenAI’s newly appointed head of strategic futures as a “supreme village idiot.” These remarks highlight growing tensions within the Trump administration regarding the implications of Chinese advancements in AI. The controversy stems from the emergence of Kimi, a free, open-source AI model developed by Moonshot. Kimi K3 reportedly matches or exceeds the capabilities of paid models from OpenAI and Anthropic, prompting concerns among U.S. policymakers and industry leaders. The availability of high-quality, cost-free alternatives from China threatens the business models of American AI companies, which rely heavily on subscription fees and premium services. This economic pressure is intensifying political friction, as the Trump administration grapples with how to balance national interests against the global AI landscape. At the same time, Anthropic has reached a landmark agreement with copyright holders, resulting in a $1.5 billion settlement. The company admitted to using pirated works to train its AI model, Claude, and agreed to pay the largest known copyright payout in history. While the ruling represents a significant legal precedent, many creators and authors remain skeptical, arguing that the settlement does little to address broader concerns about AI’s impact on intellectual property rights. Some experts warn that heightened fears around AI copyright could stifle innovation and creativity in the long term. Separately, the Trump administration is considering measures to restrict the use of Chinese AI models, citing national security and economic competition as key motivations. The launch of Kimi K3 has reignited discussions about imposing bans or limitations on Chinese-developed AI technologies. However, there is no consensus within the administration on the specifics of such policies. Some officials advocate for stricter regulations, while others argue that such actions could hinder technological progress and international collaboration. Meanwhile, China is exploring ways to tighten control over the export of AI-related technologies, aiming to prevent Western entities from accessing its cutting-edge innovations. Reports indicate that Beijing has engaged in discussions with technology firms about implementing export restrictions, reflecting a broader strategy to safeguard its intellectual property and maintain competitive advantages. These moves underscore the increasing geopolitical stakes surrounding AI development and deployment. In another development, Google is advancing its efforts to enhance AI performance through hardware improvements. The company is developing a new chip, dubbed “Frozen V2,” designed to run its Gemini models more efficiently. If successful, the chip could significantly boost the speed and scalability of AI applications, potentially giving Google a competitive edge in the rapidly evolving AI market. The announcement has already contributed to a surge in Alphabet’s stock price, signaling investor confidence in the company’s future prospects. On a different front, law enforcement agencies in New Orleans are exploring the integration of armed drones into their operations. A draft drone manual suggests that quadcopters could be equipped with weapons, raising ethical and legal questions about the use of force in policing. Such developments highlight the expanding role of AI and automation in public safety, even as they spark debates about accountability and oversight. These diverse developments illustrate the complex and multifaceted nature of the AI landscape, encompassing both technological innovation and regulatory challenges. As the field continues to evolve, stakeholders across the globe will need to navigate these issues with careful consideration of their societal and economic impacts.