Matching stories Earlier stories
Google releases a PyTorch developer tool to move apps off Nvidia GPUs
Google follows Amazon with a developer tool that helps PyTorch users shift their apps off Nvidia GPUs.
OpenAI lets ChatGPT users lock individual chats
The lock is meant for conversations users want kept private when they hand their device to someone else to show them a different chat.
Eric Schmidt reportedly holds about half the Anthropic stake of CEO Dario Amodei
Bloomberg reports that former Google CEO Eric Schmidt owns roughly a 1% stake in Anthropic, about half of what Dario Amodei holds.
Meta ordered to name executives who approved torrenting books for AI training
A magistrate judge in copyright suits by book authors and Entrepreneur Media ordered Meta to identify decision-makers behind downloading books from shadow libraries like Anna's Archive and LibGen, per TorrentFreak as cited by Ars Technica.
Meta's Muse AI companion passes 10 million U.S. downloads in 40 days
Muse reached more than 10 million U.S. downloads in its first 40 days, a milestone reached by only 11 other apps, including Threads. New data shows it retains users at higher rates than rival AI apps.
Gradium releases Phonon, open-source on-device text-to-speech models
Gradium released Phonon, two small TTS models of 31M and 40M parameters, with a fully open-source inference stack targeting CPUs, GPUs and NPUs across 10+ backends, mostly written in Rust.
Mistral launches Voxtral 3 speech models and a voice-agent SDK
Transcription costs $0.0035 to $0.008 per minute and text to speech $13 per million characters. Mistral has not said whether weights are out.
Spotify says it removed 50 million AI slop and spam tracks from recommendations
Spotify says it has removed more than 50 million songs considered AI-generated slop or spam from user recommendations over the past year and will highlight more human-made music. It has not mandated labels on all AI tracks but launched an "AI Persona" badge.
Adobe adds one-tap Open Eyes feature to Lightroom mobile
The generative AI Open Eyes Quick Action appears automatically when the app detects closed eyes in a photo, on iOS, iPadOS and Android. The feature is available in the mobile version of Lightroom.
Eka Robotics valued at $1.45 billion on physical AI robotic arm model
Eka Robotics is reported to be worth $1.45 billion, with experts praising its robotic arm for picking up a wide range of objects smoothly and human-like.
Meta suspends Muse sharing feature after researchers expose shared conversations
Security firm Claroc enumerated 6,800 shared Muse conversation links and found sensitive personal and business data in a few hundred of them. Meta says it is suspending sharing while it reviews the design and considers changes.
Army Cyber Command develops ECHOTRIBBLE, a worm-like AI agent program
According to internal documents, the program is designed to spread across networks, including air-gapped systems, and neutralize targets with minimal human oversight. Experts warn of collateral damage.
Google and Synaptics unveil AI chips for wearables and industrial devices
The chips are built on Google's open-source Coral technology and target wearables and industrial devices.
Cloudflare releases Clairic, a model that extracts data from documents
Cloudflare says Clairic pulls data from documents faster and more cheaply than large language models.
Otter.ai raises $320 million from General Catalyst's Customer Value Fund
The meeting note-taking startup says the funding comes from General Catalyst's CVF, which provides capital without diluting equity and earns returns based on marketing spend results. Otter.ai plans to use it for enterprise growth, go-to-market, international expansion and product innovation.
OpenAI publishes case study on franchisee cutting cash reconciliation time with ChatGPT Work
An independent Schlotzsky's and Cinnabon franchisee reportedly cut menu repricing time by 95% using ChatGPT Work and built a workflow that could free up $100K in cash flow.
DeepSeek open-sources DeepInfer, the serving stack it runs in production
It ships under Apache 2.0 and for now supports only DeepSeek-V4.1-Flash in offline mode. It leads Hacker News with 1,862 points.
OpenAI's GPT-6.1 Sol powers Terminal X's Eval Agent, matching senior investor judgment 88.6%
Terminal X's Eval Agent grades outputs against a 54-criterion rubric and catches 25% more deviances from a fund's mandate. The figures come from OpenAI's customer story.
OpenAI acquires Orchestra Bio to expand AI into life sciences R&D
The acquisition brings AI-powered operations to life sciences research and development, helping biopharma teams accelerate drug development.
US bans DeepSeek technology across federal civilian agencies
The US Department of Homeland Security has barred federal civilian agencies from using AI technology from Chinese lab DeepSeek.
fal launches FLUX 3 Max video generation model
5. The claim comes from the company and no independent benchmark is cited.
CoreWeave's bonds fall into distressed territory as shares drop over 20 percent
CoreWeave's bonds slipped into distressed territory, with yields about 1,000 basis points above government bonds. Its shares are down more than 20 percent since the prior Tuesday's close, and other AI stocks saw some spillover.
MiniMax M3.1 Flash beats DeepSeek and Opus 5 on a business benchmark at lower cost
The model reportedly leads on a key business benchmark at a lower price, but shows a high hallucination rate.
Hackers breach OpenCode AI coding tool and steal user API keys
Attackers got into part of the open source coding agent's online platform, stole some users' API keys, and ran up charges on their AI accounts. AI tools the team consulted refused to analyze the malicious code.
Tolan raises $30M for its AI companion app
The AI companion app Tolan has raised another $30 million after reaching more than 250,000 active users.
Kling AI releases Kling 4.0 video generator with stereo audio and lip sync
Kling 4.0 supports clips up to 30 seconds, ten keyframe images as input, and prompts of up to 8,000 tokens, plus a lighter Flash variant. Comparisons with Wan 3.0 and Seedance 2.5 show it holding its own without clearly pulling ahead.
MiniMax M3.1 Flash claims nearly 3x faster inference at $0.15 per million input tokens
Output costs $0.60 per million. Weights come later, and a larger M3.1 Pro is planned.
Disney hires most of Character.AI staff and may acquire its technology after bankruptcy
Court filings show Character.AI filed for bankruptcy as part of a deal letting Disney hire most staff and potentially acquire much of its technology, leaving the remaining company to handle lawsuits and regulator investigations.
FLUX 3 Max renders a 10-second video in about 12 seconds
Its tie for first on image to video comes from a preliminary Artificial Analysis arena.