Legora raises $550 million to compete with Harvey in legal AI

Anthropic, one of the companies whose artificial intelligence products have become popular worldwide alongside OpenAI, suffered significant financial losses in the first half of this year.

According to a report published by The Wall Street Journal on Tuesday, Anthropic recorded a loss exceeding 80 billion US dollars for the first half of this year. When converted into Indian currency, this amount is approximately 75,76888 crore rupees.

The company also reported that it may spend $750 billion next year to improve computing power.

The report indicates that last year the company's losses were about $13 billion, but there has been a sharp increase in losses during the current period. This jump is due to accounting expenses of $72 billion, which are directly related to the change in the value of financial instruments issued by Anthropic to raise funds, according to the company's IPO prospectus draft.

Anthropic's losses demonstrate how costly new technologies have become in the current situation. Currently, most companies provide their AI services for free.

Nevertheless, The Wall Street Journal also noted that Anthropic's revenue increased over the six-month period, reaching $16.3 billion. A year earlier, this revenue figure was about $1.1 billion.

Mueon Corporation has successfully raised $65 million in a Series A funding round to develop its modular computing platform. The round was led by Maverick Silicon, with participation from Intel Capital and other investors.

Chief Executive Officer and co-founder Wilfred Gomes stated that the funding will enable the company to bring integrated computing and memory systems to market. Additional investors included Geodesic Alliances Fund, Oregon Venture Fund, M-Ventures, Wellington Capital, IQT, and Defined.

As a result, Mueon's total capital raised now exceeds $80 million. Previously, the company secured $15.5 million in seed funding in 2025.

Mueon develops semiconductor systems designed for high-performance computing and artificial intelligence. The company's technology aims to meet the growing infrastructure needs of data centers. The new funds will be used to expand engineering and operations teams.

The financing will also help advance the Cubelets technology from a platform development stage to a broader validation system. The company plans to prepare this technology for production while deepening customer engagement. Gomes added that Mueon's architecture is designed to overcome existing rack-level limitations.

Mueon's core technology, known as Cubelets, consists of modular building blocks for three-dimensional computing and memory. This architecture integrates compute power, memory, interconnects, power delivery, and thermal management. Unlike traditional data center systems that treat these components as separate design tasks, Mueon designs them as parts of a unified system.

This approach is intended to reduce data transmission distances and lower energy consumption. The company aims to achieve a tenfold improvement in energy efficiency and cost within a single rack. Furthermore, the system architecture is designed to scale to megawatt-level density if required by applications, although these figures are company goals and have not yet undergone independent verification.

Key aspects of the work include silicon architecture and design development, advanced integration, and software orchestration and ecosystem validation. These elements cover hardware architecture, thermal design, power optimization, and electrical parameters, as well as compatibility definition with existing AI frameworks.

The investment community also gained representation on Mueon's board of directors. Manish Muthal, Senior Managing Director at Maverick Silicon, will join the board. Srini Ananth, Managing Director at Intel Capital, will also serve as a director. Both investors emphasized the company's systemic approach to data center design.

Industry partners have also shown interest in the platform's potential. d-Matrix sees opportunities in combining modular infrastructure with in-memory computing. Ampere Computing, now part of SoftBank, is also collaborating with Mueon, focusing on mitigating technological risks as the platform moves toward production. Mueon is also working with technology and manufacturing partners.

The increasing demand for AI workloads continues to strain the power and cooling capacity of data centers. Operators require infrastructure capable of scaling with expanding workloads without proportionally increasing overhead. Mueon's methodology seeks to solve this problem through modular system integration.

The next phase of funding will focus on growing engineering capabilities, engagement, and production readiness. It is necessary to demonstrate that Cubelets can achieve the expected efficiency gains at scale. Successful validation could help establish Mueon's architecture as an alternative for future AI infrastructure.