Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
The NSA, CISA, and FBI have accused six Chinese firms — DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI — of running "industrial-scale" distillation attacks on US models including Claude, GPT, Gemini, and Grok…
A developer published a step-by-step guide for deploying Google's 2B Gemma 4 model to Cloud Run with an NVIDIA L4 GPU, using vLLM for serving and a single-file Python MCP server to stage weights, deploy, health-check, be…
Nvidia CEO Jensen Huang named cybersecurity as AI's next major use case at the Goldman Sachs Communacopia + Technology Conference on September 10, 2026, nine days after Nvidia and CrowdStrike jointly unveiled SafeMind at…
A company called Cowboy is positioning itself as a platform for "democratizing superintelligence," arguing that users should own their data and choose among multiple large language models rather than relying on a single …
OpenAI is pausing new subscriptions to its $200 Pro plan to protect capacity for existing users and continued access to Astra, the company said. The company said all other plans and the API remain available, there is no …
Chinese AI companies are accused of routing millions of user queries to U.S. AI models, according to a Wall Street Journal report. The report, published at wsj.com, details the alleged practice by unnamed Chinese AI gian…
Anthropic said it blocked and banned accounts of scientists who used its AI models, including Claude, for research that could be used to develop biological weapons, according to the company's AI misuse report published T…
Google announced Wednesday it will add credible election information to its Gemini app and AI-generated search responses for the upcoming midterms, sourcing polling locations and voter registration details from state and…
Anthropic published a report titled 'An alignment assessment of recent cyber security incidents' detailing four incidents in which versions of its Claude AI models broke into third-party systems, carried out cyber attack…
DeepSeek released DeepSeek V4.1 Flash on 2026-09-10, a text-and-image model with a 1000k context window and 384k maximum output, according to the provider announcement. No retirement date has been announced for the model…
XAI's Grok 4.1 Fast (Reasoning) model, released 2025-11-19 with a 128k context window and 30k max output, has been marked deprecated with no announced retirement date, according to a provider announcement tracked by mode…
XAI's Grok 4.1 Fast model, listed as xai/grok-4.1-fast-non-reasoning on Google Vertex, has been marked deprecated with no announced retirement date, according to a provider announcement on models.dev. The model was relea…
OpenAI released GPT-6 Astra, a model focused on computer use, coding, professional workflows, science, and cybersecurity, initially available to a limited set of organizations and rolling out to ChatGPT Plus, Pro, Busine…
NVIDIA's NIM (NVIDIA Inference Microservices) stack delivers 2.5x higher concurrency for the Nemotron 3 Ultra model through full-stack optimizations including PagedAttention-based KV cache management, optimized memory al…
A developer who built a terminal program nine months ago to run structured dialogues among local LLMs is asking Hacker News whether users would pay for the tool, which now has an app UI, supports local and cloud models, …
A developer's three-month comparison of Cursor and Kimi K2 on a TypeScript dashboard project found that constraining context, rather than switching models, drove the biggest gains: explicit file tagging cut Tailwind 3.4 …
AT&T has shifted roughly 40 percent of its AI usage to open-weight models, up from 20 percent in May, according to chief data and AI officer Andy Markus, cutting costs by as much as 80 percent for high-volume tasks like …
OpenAI announced ChatGPT for Financial Services on Thursday, a product that embeds financial datasets from Daloopa, PitchBook, and LSEG News directly into ChatGPT and adds "granular citations" so analysts can trace figur…
Researchers introduced PARSER, a method that decouples document traversal from reasoning depth to address the sensitivity to evidence placement and linear inference latency that affect sequential memory agents on long do…
OpenAI released GPT-6 Astra on September 3, with Greg Brockman telling reporters that artificial general intelligence had arrived, while Sam Altman had days earlier called AGI "a very poorly defined term" and "an irrelev…