2026
10,721 articles · 136 days · 10 categories · avg 79 per day
OpenAI launches GPT-5.6 and a new model family with security improvements
On July 9, 2026, OpenAI unveiled a new family of language models centered on GPT-5.6. The update promises improvements in cybersecurity and other areas of AI…
Fastest-Growing AI Startups Are Accelerating Their Revenue Growth
Among fast-growing AI startups, a group of leaders is increasing revenue much faster than its competitors and the industry average.
Henry Schein One rolls out AI X-ray checks across 10,000 clinics
Henry Schein One launched Image Verify on Amazon SageMaker AI, a real-time quality check for dental X-rays, and deployed it to 10,000 locations within a few months.
UN AI Summit: Can Regulation Keep Up With the Technology?
Sunrun will offer customers payments to host AI compute nodes in their homes
Australian Payments Plus speeds up payments with ChatGPT Enterprise and Codex
How to Use ChatGPT in Schools: OpenAI's Teacher Training Program
OpenAI criticizes SWE-Bench Pro over unreliable scoring
OpenAI analyzed the popular SWE-Bench Pro benchmark and found problems with how reliably it measures AI models' coding ability.
OpenAI launches GPT-5.6 family with three models and an ultra mode
OpenAI unveiled GPT-5.6 Sol, Terra, and Luna, with gains in coding, agentic work, design, science, and API efficiency.
OpenAI unveils ChatGPT Work, an agent for hours-long projects
Deutsche Telekom and OpenAI Bring AI to Telecommunications
AegisDx: AI framework for safer diagnosis outscores GPT-5 in physician ratings
Harness Engineering: Moving LLM Agent Guarantees from Prompts into Code
Self-evolving LLM agents: compiling SOPs into tools cuts latency by 42%
Researchers show that precompiling routine SOP steps into verified tools cuts agent latency by 42% and error rates by 53% in a real production deployment.
Graph neural network recognizes hand gestures in 48 ms with 99% accuracy for prosthetics and AR
Researchers built a graph neural network that classifies hand gestures from forearm muscle signals in 48 ms with 99% accuracy, a new benchmark for prosthetics and…
Qwen3 language models raise task difficulty easily but struggle to simplify it
Researchers tested Qwen3-Next-80B and Qwen3-Coder-Next on adapting programming tasks along Bloom's taxonomy and found a consistent asymmetry between making tasks…
New MiniLM-based method better spots queries outside a chatbot's knowledge scope
Debiasing NLP Models Creates New Stereotypes for Other Demographic Groups
Researchers propose "alignment plausibility" as a safety standard for LLMs in healthcare
A review of 18 medical LLMs: specialist models diagnose more accurately than general-purpose ones
EspanStereo: A New Dataset Exposes LLM Stereotypes Across Spanish-Speaking Cultures
Researchers have built EspanStereo, the first Spanish-language dataset of cultural stereotypes for language models, covering Spain and Latin America.
DeepSearch-Evolve: a 9B-parameter web agent trains without a teacher and scores 61% on GAIA
The DeepSearch-Evolve self-distillation framework trains web agents on their own experience across 420,000 tasks, with no distillation from powerful closed models.
LLMs prove theorems but don't discover mathematics: researchers outline the next step
New AI Tool Combines GTAP and APSIM to Assess Shocks in Agricultural Supply Chains
LSTM beats Random Forest and Naive Bayes at tweet sentiment analysis with 0.92 ROC-AUC
Context Graph: Researchers Teach Enterprise AI Agents to Work Without Being Asked
Eternis-Forecaster 8B: Activation Probes Beat Chains of Thought and Save Up to 47% of Tokens
Researchers show that probes of intermediate LLM activations reveal a model's real forecasts more accurately than chain-of-thought, and can save 30–47% of tokens…
AI for documents: Datalab Lift vs. the competition
Datalab compared its 9B Lift tool with NuExtract3, LlamaExtract, Marker, and Docling, and explained when a schema-first approach wins at extracting data from PDFs.
Anthropic launches Reflect, a dashboard for analyzing how you use Claude
Anthropic has launched Reflect in beta: a dashboard showing your Claude usage history and patterns over 1–12 months, with quiet hours and tips based on 4D AI…
Managing AI Prompts in the Enterprise: Mistral Adds Version Control to Studio
Amazon closes Mechanical Turk to new customers, ending an era of crowdsourced AI
AI in education: what Forge Prep and Alpha School are selling
