Meta’s independent Oversight Board is set to expand its scope significantly. Starting with a pilot in 2026, it will begin reviewing account-level penalties, like suspensions, which have long frustrated users.
Microsoft is moving to block Exchange Web Services (EWS) access for certain low-tier licenses starting March 2026. The legacy API will be completely disabled for all Exchange Online mailboxes by October 2026, pushing users toward the Microsoft Graph API.
Uber is launching a robotaxi service in Dallas today using vehicles from startup Avride. The service will start with human safety drivers and charge standard UberX rates. This is one of over 20 autonomous vehicle partnerships Uber has forged.
Google is running a mobile-only test that integrates its conversational AI Mode directly into the search results page. This allows users to ask follow-up questions to AI Overviews without switching tabs. The move signals a major shift in how Google envisions the future of search.
ING is launching a dedicated private markets business, aiming to connect alternative capital with corporate clients. The move follows a broader banking trend but comes with significant execution risks.
Stablecoins are hitting a wall as they try to go mainstream with enterprises. The very complexity they were meant to solve is back, but now it’s scattered across a maze of crypto rails. The industry’s answer? A new kind of clearinghouse.
The MuddyWater hacking group has deployed new custom malware, including the “MuddyViper” backdoor, in a campaign from late 2024 to mid-2025. This represents a significant evolution towards more advanced, stealth-focused tradecraft for the historically noisy threat actor.
According to 9to5Mac, ChatGPT is having issues today. The web version is unresponsive for many, though the Mac app appears functional.
According to a leaked memo, OpenAI is in a “code red” crisis, shifting all resources to improve ChatGPT. The move delays planned AI agents and advertising initiatives as the company preps a new model to compete with Google’s Gemini.
A new paper argues that mimicking brain architecture in AI design is more important than massive data training. The approach could reduce the need for colossal, expensive data centers.
