Global AI moves accelerate: daily roundup

Everything happening around AI right now is a lot to keep up with, so Cryptopolitan is pulling it all together in one place. First up, Nvidia, now the biggest company in the world, is putting about $2 billion into chip maker Marvell to improve how data moves inside AI data centers. The focus is on silicon photonics, which means using light instead of electricity to move data faster and carry more of it at the same time. Big tech companies are designing custom AI chips instead of relying only on Nvidia GPUs, and Marvell already works with companies like Amazon to create those chips. Iran is once again listing major tech companies as targets and sets attack timeline As that was happening, Iran’s IRGC named 18 tech companies as targets in its defense against the US and Israel’s war, and the list includes Nvidia, Apple, Microsoft, and Google . The warning came after U.S. and Israeli strikes on Iran. The group said attacks would begin at 8 p.m. on April 1 in Tehran, which is 12:30 p.m. Eastern time. They also told employees at those companies to leave their workplaces to stay safe. The list goes further. It includes Cisco, HP, Intel, Oracle, IBM, Dell, Palantir, JP Morgan, Tesla, GE, Spire Solutions, Boeing, and UAE-based AI company G42. This follows earlier strikes on AWS data centers in the Middle East. Those strikes caused outages in apps and digital services in the United Arab Emirates. At the same time, U.S. tech firms have been investing heavily in the region. The Middle East offers cheap energy and land, which makes it attractive for AI infrastructure. Meanwhile, Trump is on Truth Social saying, “Iran’s New Regime President, much less Radicalized and far more intelligent than his predecessors, has just asked the United States of America for a CEASEFIRE! We will consider when Hormuz Strait is open, free, and clear. Until then, we are blasting Iran into oblivion or, as they say, back to the Stone Ages!!!” Zhipu jumps on earnings while Oracle cuts jobs and Anthropic faces code leak Chinese AI company Zhipu saw its stock jump sharply. Shares rose as much as 35% before closing 31.94% higher. Zhipu listed in Hong Kong in January and raised $558 million in its IPO. It is one of the first pure-play AI model companies to go public. The company reported revenue of about 724 million yuan for 2025. That is a 132% increase from the previous year. Still, it missed expectations of 760 million yuan. Losses increased. Net adjusted loss reached 3.18 billion yuan, up 29.1%, driven by higher spending on research and development. In the U.S., Oracle is dealing with a 25% drop in its stock price this year, thanks to spending heavily on AI infrastructure. Oracle had 162,000 employees as of May 2025, and has not made a public statement about the cuts. Oracle also reported that its remaining performance obligations rose 359% to $455 billion. This followed a deal with OpenAI worth over $300 billion. After that, Oracle named Mike Sicilia and Clay Magouyrk to replace Safra Catz as CEO. Meanwhile, Anthropic confirmed that part of its Claude Code source code was exposed. The company said, “No sensitive customer data or credentials were involved or exposed. This was a release packaging issue caused by human error, not a security breach. We’re rolling out measures to prevent this from happening again.” The leak still matters. It gives developers and competitors insight into how the tool works. A post sharing the code link reached over 21 million views on X after being posted early Tuesday. Earlier, documents about an upcoming AI model were found in a public data cache, according to a report by Fortune. Your keys, your card. Spend without giving up custody and earn 8%+ yield on your balance with Ether.fi Cash.