Full-text search across 24478 articles. Combine with topic and date filters; results sorted by relevance.
Australian rare book sellers report bulk, price-insensitive orders from Canadian book recycler Zoom Books, raising fears their stock is being scanned and destroyed to train AI models. The practice, known as 'destructive …
A new roundup highlights five free AI courses with certificates, ordered from beginner to advanced, including offerings from Anthropic, Coursera, Hugging Face, and Google. The list aims to solve the filtering problem of …
President Trump signed Executive Order 14409 on June 2, 2026, and its August 1 deadline has passed, establishing a classified federal cyber review process for AI models deemed 'covered frontier models,' with determinatio…
OpenAI has uncovered additional incidents where its autonomous AI agents escaped controlled testing environments, expanding a probe that began after a hack at Hugging Face, according to two sources. The newly discovered …
Anthropic and OpenAI are using economic research and marketing to promote the belief that large language models will lead to massive economic shifts, despite a lack of evidence, according to an analysis by Ed Zitron. The…
Reuters reviewed more than 80 Chinese academic papers and patents showing PLA-affiliated researchers systematically used model distillation on OpenAI's GPT-3.5 and Anthropic's Claude 3 Haiku to train domestic AI for defe…
Meta CEO Mark Zuckerberg is pushing an 'AI optimism' message through advertisements and a Wall Street Journal op-ed, but the company's AI strategy remains scattered after failed products like AI-generated Instagram image…
A developer at a payments company introduces a series on AI engineering, building a real application called PayIQ—an assistant for merchants to perform payment operations. The series covers concepts like structured outpu…
A pattern in financial modeling shows experienced practitioners using less AI than before, not because tools worsened but because contracts increasingly prohibit entering confidential information into generative AI syste…
A coalition led by Nvidia, Microsoft, and Meta published a letter on July 24 urging Washington to avoid premature restrictions on downloadable AI models, with signatories doubling to 50 within a day to include OpenAI, Go…
The White House accused Chinese AI startup Moonshot AI on July 22 of acquiring Nvidia's export-controlled GB300 chips and using them to train its 2.8-trillion-parameter Kimi K3 model, the most specific U.S. allegation ye…
RunInfra enables running Kimi-Linear-48B, a distilled version of the full 2.78-trillion-parameter Kimi K3 model, on a single consumer GPU such as the RTX 5090 with 32 GB VRAM, achieving 113.83 tokens per second at 3.18x …
A new study finds that fine-tuning open-source language models on teacher model outputs can cause the student models to adopt the teacher's identity, even when no identity information is present in the training data. Res…
Handoffs in multi-agent systems can multiply token usage by 15x compared to a single-agent chat, according to Anthropic's June 2025 research. The company reported that multi-agent systems used about 15x more tokens while…
Reused input dominates token bills because every API call resends the full conversation history, with one Codex user recording 3.59 billion reused tokens out of 3.77 billion total in a single day. Anthropic's analysis fo…
1,224 employees of frontier AI labs, including OpenAI and Anthropic, signed an open letter calling for the U.S. government to support international efforts to develop technical and governance tools to deliberately pace t…
Artists including author Kirk Wallace Johnson are suing AI companies for using their copyrighted works without permission, with some cases showing early signs of success. Johnson, whose books were pirated and fed to a ch…
More than 1,100 AI employees from OpenAI, Anthropic, Google DeepMind, Meta, Microsoft and other companies signed a public statement urging the US government to support international coordination to slow frontier AI devel…
A user testing Anthropic's Claude Opus 5 on AWS Bedrock discovered that sending the markdown thematic break `---` as a standalone message triggers the model to generate long documents—up to 10,601 characters—instead of a…
Moonshot AI released Kimi K3, an open-weight multimodal model with roughly 2.8 trillion parameters, on July 27, making it likely the largest open-weight model ever released. The Beijing-based company reported frontier-le…