OpenAI and the Gates Foundation have partnered for a workshop in Asia, aiming to empower disaster response teams. The initiative focuses on the practical application of artificial intelligence to enhance the effectiveness of interventions in critical scenarios, addressing challenges related to technology deployment in complex environments and sensitive data management.
Phison and Intel are collaborating to develop a hybrid AI approach, aiming to meet the growing demand for processing capabilities in China. This strategy balances on-premise and cloud resources to optimize costs, latency, and data sovereignty, offering a scalable and sustainable solution for rapidly evolving AI infrastructures.
MAGNET is a decentralized system for autonomous generation, training, and serving of domain-expert LLMs on commodity hardware. It integrates autoresearch, BitNet b1.58 training for CPU-native inference without GPUs, and distributed merging. It tracks contributions on-chain. Its capabilities have been validated in areas like video safety classification, crypto prediction, and BitNet optimization, demonstrating significant potential for autonomous and efficient AI deployments.
OpenAI has discontinued Sora, its AI video-generation tool, just six months after its public release. The decision immediately raised questions, particularly regarding the feature that allowed users to upload their faces. It is speculated that the move might have concealed an elaborate large-scale data acquisition strategy, a critical aspect for data sovereignty and privacy.
As Large Language Models (LLMs) continue to evolve, the true challenge for AI adoption in real-world scenarios stems from managing physical data. Joyce Wen from Aurotek highlights how the collection, processing, and sovereignty of this data represent a significant hurdle, pushing companies towards on-premise deployment solutions to address infrastructural and compliance complexities.
Swancor has introduced artificial intelligence-powered robots, distinguished by their full recyclability and the use of advanced composite materials. This innovation aims to facilitate large-scale industrial adoption, proposing solutions that integrate AI capabilities with principles of environmental sustainability. The initiative underscores the importance of considering the complete lifecycle of robotic technologies, from production to disassembly, for reduced ecological impact and enhanced operational efficiency in on-premise settings.
NeurIPS's recent decision to exclude OFAC-listed institutions from its activities has sparked a heated debate. This move, aimed at ensuring compliance with US regulations, raises significant questions about the nature of international scientific collaboration and access to AI research, with potential repercussions for deployment strategies and data sovereignty for global organizations.
DIGITIMES Research's analysis highlights a transition in the AI computing landscape: the focus is increasingly shifting towards inference. This change, presented at AI EXPO 2026 by Jim Hsiao, senior analyst, is redefining the challenges and bottlenecks within data centers, demanding new strategies for resource optimization and workload management.
Taiwan is intensifying efforts to integrate into the space sector's supply chains, focusing on a strategic collaboration with the United Kingdom. This move reflects the ambition to diversify its technological and production capabilities, exploring new opportunities in a high-innovation sector. The initiative could have implications for technological sovereignty and the resilience of critical infrastructures.
A consortium in Tokyo is testing the installation of data centers under railway overpasses. This solution presents significant challenges due to vibrations and thermal variations induced by passing trains, requiring innovative solutions to ensure stability and reliability.
A Japanese company has developed an optical fiber that can quadruple the data transmission capacity of existing undersea cables. The new fiber, while maintaining the same diameter, promises to significantly improve global network infrastructures without requiring major structural changes.
Last week saw a surge of investments in Europe, with a particular focus on infrastructural layers. Funding spanned diverse sectors such as semiconductor physics, orbital logistics, defense systems, and artificial intelligence, signaling a strong interest in foundational technologies.
Nginx 1.29.7, the newest mainline version, introduces Multipath TCP support. Released alongside the stable version 1.28.3, this release also addresses vulnerabilities, including a buffer overflow.
Meta partners with Entergy to build seven new natural gas power plants. The goal is to deliver 7 gigawatts of power to its planned AI data center in Louisiana, ensuring sufficient energy for compute-intensive operations.
The expansion of AI infrastructure into the real world is meeting resistance. An AI company offered an 82-year-old woman $26 million to build a data center on her land, but she refused. Tensions are rising regarding the territorial and social impact of AI deployments.
Microsoft is tightening requirements for Windows kernel drivers, excluding those not compliant with the Windows Hardware Compatibility Program (WHCP) to enhance operating system security. This move aims to reduce vulnerabilities stemming from unverified code.
A coalition of European enterprises has launched Euro-Office, an open-source office suite compatible with Microsoft formats. The goal is to provide a reliable and sovereign solution for public administrations, businesses, and educational institutions, reducing dependence on non-European platforms and ensuring data sovereignty.
According to Trump administration officials, SMIC allegedly transferred chip manufacturing tools to the Iranian military. The transfer would also include technical training on SMIC's semiconductor technology.
Ysios Capital, Spain's largest life sciences VC, has launched a new fund of €100 million to build biotech companies from university and research center spinouts.
The use of large language models (LLMs) in industrial environments, directly in factories, is emerging as a high-value, yet under-discussed application. The use of on-premise solutions, as demonstrated by some plant engineers, overcomes legal and confidentiality restrictions that would prevent the use of external cloud-based APIs for production data analysis.