Apple filed a lawsuit against OpenAI on July 11 in the U.S. District Court for Northern California, alleging theft of trade secrets. The company claims that the hiring of over 400 former Apple employees by OpenAI was not merely a case of job transitions but part of an organized effort to extract confidential technology related to semiconductor design, on-device AI, and hardware design. OpenAI has been pursuing its own hardware business since acquiring Jony Ive's design firm and recently recruited Noam Shazeer from Google DeepMind.
If Apple gains an advantage in initial proceedings or extensive discovery, it could impact hiring practices across frontier AI companies that have absorbed significant talent from big tech firms.
OpenAI prepares for IPO with a valuation of $730 billion
OpenAI is reportedly preparing a confidential IPO application with Goldman Sachs and Morgan Stanley, aiming for a valuation of approximately $730 billion as early as September. If successful, this would mark the largest technology IPO in history.
However, the market environment poses challenges. According to Fortune, Anthropic has surpassed OpenAI in annualized revenue, reporting about $47 billion compared to OpenAI's projected $25 billion to $33 billion for 2026. Anthropic's Claude product alone saw its revenue increase from an annualized $1 billion at the end of 2025 to over $2.5 billion in February. The Apple lawsuit and Anthropic's revenue lead are seen as significant variables affecting the IPO timeline.
Meta plans to double computing capacity by 2027
According to internal memos, Meta plans to double its total computing capacity by 2027 through long-term contracts with suppliers, including Samsung. On the same day, the company announced a $10 billion investment to build a 1-gigawatt data center in Alberta, Canada, which led to a more than 7% surge in its stock price. Meta's custom AI chip, Iris, co-designed with Broadcom and manufactured by TSMC, is set to enter mass production in September. This chip will handle internal inference workloads, complementing rather than replacing NVIDIA GPUs.
U.S. allows unlicensed AI chip exports to UAE
The U.S. Department of Commerce has reclassified the United Arab Emirates (UAE) from an export control country, allowing unlicensed exports of advanced AI chips. Companies like Amazon, Apple, and xAI, which are building large-scale infrastructure in the UAE, will benefit directly. This move aligns with the Biden administration's summer strategy of easing AI trade regulations with allied nations while maintaining barriers against China, potentially diversifying frontier-level learning infrastructure beyond U.S. borders.
Federal Reserve establishes first AI task force amid conflict of interest concerns
The Federal Reserve has created a task force to study the impact of AI on employment, productivity, and monetary policy, appointing a16z co-founder Marc Andreessen as co-chair. This marks the first official body within the Fed to address the economic effects of AI. However, concerns have been raised about potential conflicts of interest, given that Andreessen has invested billions in AI companies and is now leading central bank analysis of the technology.
Humanoid robot companies rush to go public
In the past week, three humanoid robot companies have initiated public listing procedures. Agility Robotics has applied for a SPAC listing with a valuation of $2.5 billion, while China's Unitree has received IPO approval on the Shanghai Stock Exchange. Tesla has begun converting its existing production lines into a dedicated factory for its Optimus robot. However, all three companies have yet to demonstrate unit economics—such as component costs, on-site reliability, and actual labor replacement rates—at scale, making the post-IPO disclosure of deployment and margin figures critical for market validation.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.