News Archive
Pick a date to read that day's stories.
Days with a blue dot have stories.
- 01 of 10FUNDING
Mistral eyes $3.5B raise at $23B valuation
French AI lab Mistral is raising three point five billion dollars in fresh capital at a twenty three billion dollar valuation. Backers include ASML, Nvidia, Andreessen Horowitz and France's BPI. It would nearly double the firm's worth in a year.
OUR TAKEEurope's frontier bet keeps doubling like a stubborn souffle.
Read more → - 02 of 10ENTERPRISE
OpenAI launches Presence for business voice agents
OpenAI unveiled Presence, a platform for enterprises to build and run realtime voice and chat agents. It handles routing, memory and human handoff. Access opens for eligible business customers this week.
OUR TAKEThe pitch is a call centre that never has a bad Monday.
Read more → - 03 of 10HEALTH
HHS joins US mission to speed drug discovery with AI
The US Department of Health and Human Services signed on to the Genesis Mission, pledging labs and data to help AI models chase chronic disease treatments. It follows the White House's five billion dollar science push.
OUR TAKEIf it shaves years off a drug trial, that is worth every dollar.
Read more → - 04 of 10ROBOTICS
Applied Intuition unveils Dana for physical AI
Applied Intuition launched Dana, an agentic platform for building, testing and running AI systems for real world machines. Target customers span cars, trucks, warehouse robots and factory kit. It builds on the firm's decade of simulation work.
OUR TAKEThe playbook that trained self-driving cars is coming for every moving box.
Read more → - 05 of 10RESEARCH
Tencent's Hyra-1.0 agent runs its own science
Tencent's Hunyuan team released Hyra-1.0, a research agent that can pick a topic, run experiments and rewrite its own code between rounds. Early demos cover drug design and materials.
OUR TAKEThe lab assistant that never asks for a coffee break.
Read more → - 06 of 10POLICY
WHO sets ethics rules for AI in health research
The World Health Organization published new guidance for ethics committees reviewing AI use in health studies. It covers consent, bias checks and how to audit models after deployment. The rules apply worldwide.
OUR TAKEFinally a rulebook clinics can actually hand a review board.
Read more → - 07 of 10SECURITY
Cisco open-sources Antares bug-hunter models
Cisco released two small open-weight AI models that hunt for software vulnerabilities. Companies can run them locally to scan their own code repeatedly without shipping source outside. The models are free to download.
OUR TAKEA quiet gift for every underfunded security team on the planet.
Read more → - 08 of 10FUNDING
Glow raises $180M for AI endpoint security
Cyber startup Glow emerged from stealth with a one hundred eighty million dollar Series A and a one point two billion dollar valuation. Founded by former Meta and Snowflake leaders, it uses AI to protect laptops and phones from novel attacks.
OUR TAKEAntivirus grew up, got a computer science degree and moved to San Francisco.
Read more → - 09 of 10ROBOTICS
UK's Humanoid raises £114M for factory robots
London-based Humanoid closed a one hundred fifty two million dollar Series A at a one point three five billion dollar valuation. The round is Europe's largest for a humanoid-first firm. It plans warehouse and light industry pilots first.
OUR TAKEBritain finally has a robot startup people can actually name.
Read more → - 10 of 10HARDWARE
Nvidia ships Cosmos 3 Edge for on-device AI
Nvidia released Cosmos 3 Edge, a set of small models built to run on factory, hospital and warehouse machines. They predict how a scene will change and how an action will affect the world. Weights are on Hugging Face.
OUR TAKEThe brains for robots that cannot phone home every second.
Read more →