Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
SWIRL 5 introduces a source-grounded hallucination check that runs after generation to catch confident wrong answers with misleading citations. The system segments claims, verifies them against source documents, and flag…
Google researchers have developed the Scalable Cluster Termination System (S-CTS), a multimodal defense that uses LLMs enhanced with LoRA and APO to detect and terminate coordinated bot-nets producing AI-generated spam o…
AI startups should shift from selling productivity-enhancing software to selling completed work products, opening new verticals and pricing models. EvenUp exemplifies this by selling AI-generated demand packages to perso…
The Trump administration lifted export controls on Anthropic's AI models, enabling the company to restore access to Claude Fable 5 and Mythos 5. The decision marks a shift in US AI policy and benefits Anthropic's global …
Anthropic suspended access to its new Fable 5 and Mythos 5 models three days after launch on June 12, 2026, citing unspecified issues, and retired Claude Sonnet 4 and Claude Opus 4 on June 15, breaking any requests still…
Rising token costs are driving companies to abandon tokenmaxxing strategies in favor of open weight models, which offer greater efficiency and lower computational expenses. Alex Karp noted the shift as businesses priorit…
The U.S. Department of Commerce lifted export controls on Anthropic's Claude Fable 5 and Mythos 5 models, reversing restrictions imposed in June over national-security concerns. Commerce Secretary Howard Lutnick informed…
The Trump administration lifted restrictions on Anthropic's Claude Fable 5 and Mythos 5 AI models after a cybersecurity alarm, ending a weekslong ban. The Commerce Department had blocked foreign nationals from using the …
Anthropic suspended Fable 5 and Mythos 5 due to an inability to verify nationality in real time after new US export controls, highlighting that frontier-model releases now depend on identity checks and access enforcement…
Meta unveiled Brain2Qwerty v2, a non-invasive AI system that decodes brain signals into typed sentences with 61% average word accuracy, up from 48% last year. The system uses a magnetoencephalography scanner and a large …
The US government lifted export restrictions on Anthropic's Fable 5 AI model after the company agreed to collaborate with tech giants on security measures, allowing global availability to resume. The restrictions were in…
A developer shares seven key lessons learned from building modern AI agents with Neo4j, GraphRAG, Aura Agents, and LLM Mesh. The lessons cover graph databases, persistent memory, GraphRAG vs. traditional RAG, multi-LLM o…
SelectPdf has released a free, open-source (MIT) set of Agent Skills called SelectPdf AI Tools that prevents AI coding assistants from hallucinating PDF API methods. The tools provide accurate, version-matched knowledge …
Australian startup Springboards has developed an LLM called Flint that generates more diverse responses than mainstream models like ChatGPT and Claude, addressing the issue of AI groupthink. The company's approach embrac…
At the AI Engineer conference in San Francisco, a developer observed that the industry is moving past simple chat interfaces and treating large language models like central processing units within an LLM Operating System…
A developer added a synchronous 'Generate' button to a content dashboard, replacing an async job queue that required manual server intervention. The endpoint POST /api/drafts/generate?platform= runs the generator synchro…
Telnyx published a code example showing how to build a production-ready AI chat endpoint in Node.js using the Telnyx AI Assistants API. The 80-line Express app exposes a /chat endpoint that sends messages to a configured…
A developer at the Agentic AI Foundation proposes a rubric for deciding whether to expose internal tools as an MCP server or a CLI for agent use. The rubric emphasizes that CLIs are best for human-driven, terminal-native…
Meta hired hundreds of contractors to pose as teenagers and send over 45,000 prompts about suicide, sex, drugs, and eating disorders to rival AI chatbots including OpenAI's ChatGPT, Google's Gemini, and Character.AI, acc…
Altus Schools, a San Diego charter network, spent $500,000 on two Ameca humanoid robots from Engineered Arts, integrating them with ChatGPT to serve as teaching assistants for struggling students. Critics question the co…