Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Anthropic published a policy framework urging governments to regulate the most powerful AI models, including granting legal authority to block dangerous deployments. The company cited rapid AI capability growth, such as …
A developer argues that embedded SDKs for AI agent reliability outperform proxy gateways by eliminating network latency and operational overhead. The comparison shows embedded SDKs add ~0ms latency versus 30-200ms for ga…
An analysis of 10,000 production LLM calls reveals a 5-15% first-attempt failure rate, with timeouts, rate limits, and schema violations being the most common issues. The NeuralBridge project proposes a self-healing appr…
NeuralBridge, an embedded SDK for LLM-powered AI agents, self-heals from API failures with an 84.1% auto-recovery rate. The open-source SDK recognizes over 280 fault patterns and employs 30+ recovery strategies, diagnosi…
The US Commerce Department ordered Anthropic to suspend access to its frontier AI models Fable 5 and Mythos 5 for all foreign nationals, citing a potential jailbreak. Anthropic complied but disputed the rationale, while …
President Trump banned non-American nationals from accessing Anthropic's newest AI models, Fable 5 and Mythos 5, citing national security concerns. Anthropic suspended the models for all users, raising alarms in the UK a…
Microsoft and OpenAI are planning to deploy hundreds of thousands of NVIDIA GB200 GPUs, alongside their custom Maia AI chips, as part of massive infrastructure expansion. Hyperscaler capital expenditure remains in the hu…
Anthropic disabled its Fable 5 and Mythos 5 models after the U.S. Commerce Department issued an export-control directive citing national security concerns, ordering the company to suspend access for foreign nationals. Th…
A developer used Anthropic's Claude Fable 5 to build a Figma clone in a single 28-minute session, adding 2,889 lines of code and passing 22 smoke tests. The US government issued an export control directive on June 12, ca…
The Trump administration ordered AI company Anthropic to block its most advanced AI models, including Mythos 5 and Fable 5, from foreign users, citing national security concerns. Anthropic complied by shutting down the m…
Moonshot AI released Kimi K2.7 Code, an open-weights model with one trillion parameters designed for programming, which undercuts GPT-5.5 and Claude Opus 4.8 by up to 12x on price per token. Although it trails behind in …
Vilius Vystartas argues that the unified global internet is being replaced by sovereign AI zones, driven by US export controls on AI chips and models. The January 2025 AI Diffusion Rule created a three-tier world, forcin…
Uber exhausted its entire 2026 AI budget by the second quarter, highlighting unsustainable costs as enterprises mix and match AI models from multiple providers. OpenAI is considering cutting token prices to compete with …
A coalition of 42 state attorneys general has opened an investigation into OpenAI, with New York Attorney General Letitia James serving a subpoena for documents on advertising, data handling, and model performance. The p…
A new distinction between chatbot dialogue and inline automation in AI writing tools has implications for student learning, according to an academic analysis. The author argues that inline automation, like Microsoft Copi…
A developer spent $788 on API calls from an AI coding agent in a single day, running 3,572 calls across four models. The breakdown reveals that 73% of calls went to the most expensive model, Fable 5, costing $617, while …
Anthropic disabled its Claude Fable 5 and Mythos 5 models for all customers on June 12, 2026, after a US government export control order citing national security concerns. The order, sent by Commerce Secretary Howard Lut…
Nvidia and Abridge are developing a healthcare-specific AI model, signaling a shift from generic AI to domain-specific models tailored to real clinical workflows. For builders, this underscores that durable AI products r…
Andrej Karpathy, a founding member of OpenAI and former Tesla AI lead, has joined Anthropic to work on its pre-training team. Karpathy announced the move on May 19, 2026, marking one of the highest-profile researcher mov…
A developer abandoned building a custom document Q&A system after struggling with fine-tuning and keyword search, then built a working prototype using Retrieval-Augmented Generation (RAG) with OpenAI embeddings and Chrom…