ASML’s newest EUV lithography unit dominates chip production, but its size and price raise eyebrows. The system spans a double‑decker bus, weighs over 150 tons and carries a $400 million price tag. Its extreme‑ultraviolet light, generated by firing lasers at molten tin droplets, etches the sub‑nanometer features needed for today’s most powerful processors. With roughly 90 % of global lithography capacity, the tool underpins the AI hardware race.

Anthropic sparked a regulatory clash when it disclosed Mythos, a coding model flagged as a cybersecurity hazard, then issued a safer variant called Fable. Within days, U.S. export controls froze distribution of both models, prompting Anthropic to revoke access. The episode highlights how quickly policymakers can intervene on AI tools that are not weapons, raising questions about the scope of future oversight.

Meta halted an internal AI training program that logged employee keystrokes after a data leak, while Nvidia unveiled a closed‑loop cooling design promising significant water savings for data centers. SpaceX readied a cargo‑shuttle test, and a UK AI‑driven law firm secured a win in a debt case. Together these stories illustrate the expanding reach and regulatory friction of emerging tech.