US reportedly delayed DeepSeek blacklist for eight months
The case shows how export controls can become trade bargaining chips even when security agencies have already made their case.
Page 22
View archive →OpenAI's LifeSciBench exposes AI's science gap
The 750-task benchmark shows leading models can help with life science, but still fall far short of independent research work.
Novo Nordisk breach puts pharma AI assets in focus
The alleged 1.3TB theft shows why models, datasets and developer credentials are now core pharmaceutical security risks.
Snap prices new Specs AR glasses at $2,195
Snap is pitching the standalone AR glasses as a see-through computer, but the price keeps them in developer territory.
DeepSeek's $7.4B round locks control with Liang and the state
The financing structure accepts outside capital while keeping voting power away from commercial investors.
Alibaba puts Qwen into robots with a three-model suite
Qwen-Robot Suite is software rather than hardware: three models for navigation, manipulation and world prediction, designed to run on robots from AgileX, Franka, Universal Robots and Unitree.
US backs SandboxAQ to find chip materials with AI
The Commerce Department is putting $500 million into SandboxAQ to search for alternatives to critical chipmaking materials, taking a non-voting equity stake and royalty rights in return.
OpenAI replays real chats to test new models
OpenAI’s Deployment Simulation uses 1.3 million consented historical conversations to predict how unreleased models will behave in production, including agentic tool-use failures.
SoftBank turns OpenAI into a patching service for Japan
SoftBank and OpenAI are selling an AI security service that scans, ranks and helps patch vulnerabilities for Japan critical infrastructure companies.
Ent raises $100 million to stop agent-era attacks early
RiskIQ veterans emerged from stealth with Ent, a security startup that wants endpoint tools to judge intent before damage is done.