DeepSeek GUI launched as a desktop application that transforms the DeepSeek AI model from a browser-based chatbot into a local workspace for coding, document writing, project planning, and task automa…
Amazon quietly added an AI-powered print-on-demand feature to its Shopping app, allowing users to generate custom T-shirt and hoodie designs by describing them to Alexa and ordering them with Prime sh…
Ideogram 4, a 9.3-billion-parameter text-to-image model from a team of former Google Brain researchers, topped the open-weight leaderboard for design quality, winning nearly half of blind typography t…
Google removed the vision and audio encoders from its Gemma 4 12B model, projecting raw image patches and audio waveforms directly into the language model's embedding space. The 12B model, which runs …
MiniMax released M3, an open-weight AI model with a 1 million token context window and native multimodal support, that outperformed frontier models in a 24-hour CUDA kernel optimization task by persis…
Anthropic confidentially submitted a draft S-1 registration statement to the U.S. Securities and Exchange Commission on Monday for a proposed initial public offering, marking the AI lab's first step t…
OpenAI announced that one of its internal reasoning models solved a math problem first posed by mathematician Paul Erdős in 1946, concerning the maximum number of unit-distance pairs among points on a…
The OpenBMB Community released MiniCPM Desk Pet v0.7.3, an open-source desktop application that runs the MiniCPM AI model locally on macOS and Windows. The floating desktop companion enables local AI …
OpenBMB released MiniCPM5-1B, a one-billion-parameter model that scores 40.42 on the AIME 2025 mathematics exam, outperforming larger competitors including LFM2.5-1.2B and Qwen3-0.6B. The model offers…
StepFun released Step 3.7 Flash, a 198-billion parameter sparse MoE model, which achieves 76.3% on SWE-Bench Verified, matching 97% of Claude Opus 4.6’s coding performance at one-ninth the cost—$0.19 …
Higgsfield, a San Francisco AI startup, falsely claimed its $500,000 AI-generated film "Hell Grind" premiered at the Cannes Film Festival. The Cannes festival confirmed the film was not part of its of…
A man received a 130-page report from data broker LexisNexis detailing six months of his and his wife's driving trips, collected without their explicit consent after he set up his car's infotainment s…
Nvidia CEO Jensen Huang announced a $500 billion investment in U.S. data centers earlier this year, earning praise from President Trump and a rollback of export controls on some chips. Huang then flew…
A single-character bug in the Starlette open-source framework, tracked as CVE-2026-48710 and nicknamed BadHost, allowed attackers to bypass authentication on millions of servers running AI agents and …
AionUi, an open-source AI cowork platform, launched version 2.1.4 with built-in agents and multi-agent automation that allows users to run multiple AI agents simultaneously for coding, research, and d…
DuckDuckGo's US app installs jumped 30.5% in one day and averaged 18.1% weekly growth after Google announced it would replace traditional search results with an AI agent, with iOS installs peaking at …
Microsoft and Uber are facing rising costs from AI coding tools as employee adoption outpaces budget projections, with Uber burning through its entire 2026 AI coding budget in four months. Nvidia's vi…
Anthropic stated Claude Mythos would remain restricted to roughly 40 organizations under Project Glasswing, but users briefly saw a "Mythos 1" preview model appear in the Claude Code interface the nex…
Alibaba’s Qwen3.7-Max autonomously optimized a kernel on unfamiliar hardware for 35 hours, making 1,158 tool calls and achieving a 10x speedup over the reference implementation. The model outperformed…
OpenAI, Nvidia, ElevenLabs, and Kakao are now using Google's SynthID watermarking system for AI-generated content, which Google says has already been applied to 100 billion images and videos plus 60,0…