■ TOP STORY

GOOGLE AVOIDS AD TECH BREAKUP, MUST SHARE TOOLS

A federal judge ruled Google doesn't have to divest its ad exchange but must interoperate its ad tech tools with rival systems. The decision reshapes the remedies in the landmark antitrust case rather than forcing a structural breakup.

► WHY IT MATTERS: This sets a precedent for tech antitrust enforcement favoring interoperability mandates over forced divestitures, affecting how regulators approach future monopoly cases.

PENTAGON DEPLOYS CHATGPT, GROK TO 3M PERSONNEL

The Pentagon launched ChatGPT Mil and Grok for Government on its GenAI.mil platform, providing 3 million military personnel with AI tools designed for warfighter operations. The move marks large-scale government adoption of commercial AI systems for defense applications.

► Military-scale AI deployment accelerates government validation of LLMs for mission-critical use, while creating new security and operational concerns around classified data handling.

WAYMO EXPANDS ROBOTAXI TO THREE NEW US CITIES

Waymo is launching autonomous taxi services in Denver, San Diego, and Tampa, expanding its commercial presence beyond existing markets. The rollout signals confidence in scaling autonomous vehicle operations across diverse urban environments.

► Waymo's multi-city expansion establishes autonomous ride-hailing as a viable business model while facing growing competition from Uber's UK launch using Wayve technology.

TRUMP ADMIN WEIGHS NEW SEMICONDUCTOR TARIFFS

The Trump administration is considering additional tariffs on imported semiconductors to incentivize domestic chip manufacturing, according to Commerce Secretary Howard Lutnick speaking at Semicon in Taiwan. The move aims to build US supply-chain resilience amid strong AI-driven demand.

► New chip tariffs will reshape semiconductor supply chains and manufacturing investment patterns, with direct cost implications for AI infrastructure and consumer electronics.

OPENAI GATES ASTRA'S CYBER POWERS TO SELECT PARTNERS

OpenAI plans a public release of its Astra AI model but will restrict advanced cybersecurity capabilities to select partners, giving them advance notice to strengthen defenses. The tiered release strategy reflects concerns about dual-use risks of sophisticated AI tools.

► Selective capability gating sets a new precedent for responsible AI deployment, but raises questions about whether partner access delays actually mitigates offensive cyber risks.