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 said Thursday that its Claude AI model is leading 26% of the company's model research and development and is involved in about 90% of that work in collaboration with humans, up from leading none in February and…
A developer built an MCP-based travel planning assistant that connects an LLM agent to real-time hotel data, enabling natural-language queries to return live hotel search, comparison, and booking results in about 4.2 sec…
A computational linguist and Hausa NLP/AI data specialist announced a Hausa LLM Healthcare & Safety Preference Evaluation Dataset on Hugging Face, developed with computational linguistics specialists and healthcare profe…
Newly unsealed documents in The New York Times-led lawsuit against OpenAI and Microsoft show Microsoft Director of Applied Science Brent Hecht internally called scraping news for AI training "an astonishing theft of unpr…
Documents unsealed in the copyright lawsuit New York Times v. Microsoft and OpenAI show Microsoft and OpenAI executives privately described their LLM training data practices as "an astonishing theft of unprecedented prop…
The U.S. administration and Anthropic have accused China's AI companies of "distillation," but a growing cohort of voices are questioning the impact of that claim, CNBC Technology reported. The dispute centers on whether…
Court documents unsealed Thursday in the New York Times' copyright suit against OpenAI and Microsoft show Microsoft employees in 2023 debated whether AI models scraping news articles amounted to "the largest theft of lab…
OpenAI's GPT-6 Astra is the strongest vision model Roboflow has tested, scoring 82.1% mAP@50 at low reasoning effort on the Roboflow Vision Evals object-detection benchmark — 5.4 points ahead of Qwen3.8 Max and 13.7 poin…
OpenAI released Astra for Law, a GPT-6 model configured for legal work, alongside 26 new ChatGPT plugins from legal tech vendors, but named only four — Harvey, iManage, Legora and Thomson Reuters — and did not provide th…
A developer's enterprise AI reality check argues that generative AI and large language models have become genuinely useful in production, but that most vendor promises of 10x productivity and massive ROI are overstated. …
Prompt optimization refines an existing prompt through specificity, structure, and iteration without touching the model itself, while prompt engineering designs a prompt from scratch, according to an article detailing fi…
Telnyx added Flex and Priority service tiers to Telnyx Inference, letting developers set a `service_tier` field on Chat Completions and Responses requests to trade latency against token cost on the same models and synchr…
PrismML released Bonsai 2 27B, a ternary-weight compression of Qwen3.8 27B that cuts model size from about 56 GB to 5.9 GB while retaining roughly 98% of the original model's benchmark score. In a hands-on test on an M4 …
A developer argues that prompt injection is only one of ten failure modes in the OWASP Top 10 for LLM Applications 2025, and that the other nine — including Excessive Agency, Improper Output Handling, and Unbounded Consu…
Enterprise security teams have built prompt-injection defenses almost entirely around text, leaving vision-language models that read invoices, screenshots, scanned PDFs and video largely unguarded, according to a 2026 CI…
Anthropic and OpenAI are exploring data center deployments of roughly 20-30 megawatts to bring AI inference capacity online faster, CNBC reported on September 18th. Anthropic has sounded out potential arrangements in the…
RubyLLM 2.0 adds video, speech, OCR, reranking, files, batches, and a shared API for provider-hosted tools to the Ruby framework, expanding on the chat, tools, agents, structured output, thinking, embeddings, images, tra…
Developer razaanstha released Ulka, an experimental browser-agent extension that runs locally in Chromium-based browsers and routes model requests through the user's own Vercel AI Gateway API key. Ulka pairs the fx orche…
A developer benchmarked TypeSafe's Jev, a System One model built for structured decisions, against Claude Sonnet 5 on a bounded Arbitrum Alignment judging gate across 102 archived submissions and 306 decisions per varian…
A paper submitted to arXiv on 17 Sep 2026 presents ALIBI, a semantic cover story attack that adds a small, non-executed read-only section containing a false security product narrative to compiled binaries, flipping 30 of…