Ukraine’s fiber-optic drones get squeezed by AI data center demand
Ukraine competes with Russia and data centers for China’s optical fiber supply.
Media reportsFollow AI as it changes.
Ukraine competes with Russia and data centers for China’s optical fiber supply.
OpenAI Academy is bringing free, hands-on ChatGPT workshops to caregivers in five U.S. cities, helping them build practical AI skills for everyday life.
Foundational precipitation and geostationary satellite datasets to observe the planet clearly.
New ThinkYoung research and lessons from OpenAI’s EMEA education partnerships on supporting teachers, AI literacy, and independent learning.
Discover how Singapore Airlines uses OpenAI models and API to enhance customer support, travel discovery, and employee productivity, doubling CSAT for its Kris chatbot.
New research from OpenAI and Upwork finds shifting demand for freelance work, new skill combinations, and larger budgets and more hours for AI-related projects.
November 9, 2026 — Claude Fable 5.2 and Mythos 5.2 arrive on Anthropic’s IPO roadshow day. How much thinking is worth paying for?
An MIT-led research team has received up to $53 million in funding from the Advanced Research Projects Agency for Health (ARPA-H) to develop safer, generative AI-driven alternatives to modern herbicides.
This paper studies differentially private selection in federated settings, proving any algorithm using standard noisy aggregation needs Ω(√d/ε) samples.
This paper was accepted at the Can We Trust AI Evaluation? Workshop at NeurIPS 2026. We evaluate how good large language model (LLM) judges are for personalized writing tasks.
State Space Models (SSMs) have emerged as a compelling alternative to Transformers, enabling sequence modeling with constant memory and linear compute.
Private Evolution (PE) is a differentially private algorithm for synthetic data generation.
Kunal Talwar's paper formalizes 'evaluability': whether black-box models can be evaluated when test samples may have been used in training.
The paper proposes weighted state souping, which learns query-dependent weights for composing cached states, substantially improving long-context performance on document QA, needle-in-a-haystack retrieval, variable tracking, and tool use.
How Curative built a hierarchy of agents on Claude to sign provider contracts about 50 times faster than its team did in 2025.
Hugging Face introduces "decision models" in Transformers with a new decision pipeline that returns probability distributions over predefined answers instead of generating text.
Hugging Face now lets organizations link to a Google Cloud account and pay through Google Cloud Marketplace, putting usage on one existing bill.
Cogent combines OpenAI models with enterprise context to help automatically remediate vulnerabilities with agents.
This analysis argues that the backlash against data centers is less a culture war than a fight over electricity, with grid connection and onsite generation provoking different kinds of opposition.
Before getting started, I wanted to thank my readers for making my new book, Co-Existence, a New York Times bestseller two weeks in a row.
"Do not ask again" prompt leads to passwords, API keys, and more being shared.
On the Latent Space podcast, Garrett Lord and team detail Handshake AI's rise to $1 billion annual revenue, covering finance, life‑science, and coding benchmarks such as ATLAS‑…
As a California state senator, Scott Wiener authored legislation aimed at curbing the risks of advanced AI.
Researchers at Queen’s University Belfast highlight risks such as context recovery attacks, rapid de‑authentication (2‑3 seconds), and horizontal anomalies for AI agents in zero‑…
ScrapeBench is introduced as the first benchmark to measure AI agents' compliance with website Terms of Service and robots.txt.