>
>AI News
Breaking developments and analysis from the AI industry
OpenAI disbands safety team after GPT-4o pullback, Codex hits 1M downloads
OpenAI has disbanded its long-standing safety team and pulled back its powerful GPT-4o model, a move that a Reddit investigation attributes to the AI’s massive operational costs, which have become unsustainable for the now product-focused company.
Read more
Apple's Siri AI overhaul hits snag, delays key features to iOS 26.4
Apple’s ambitious AI overhaul for Siri has hit a significant roadblock, forcing the company to delay its key features and resort to a piecemeal rollout that could stretch into next year, according to a report from WCCFtech.
Read more
OpenAI faces "QuitGPT" campaign urging subscription cancellations
A "QuitGPT" campaign is urging users to cancel their ChatGPT subscriptions, Discord Manual Input reports. The movement is a backlash against OpenAI's ties to the Trump administration. It follows user frustrations with the AI's performance.
Read more
Google Tests 4 Gemini 3 Pro Variants, Expands AI to Maps
Google is secretly testing four distinct variants of its next-generation Gemini 3 Pro AI model, according to a report from Mastodon Social ML Timeline, a major behind-the-scenes push that coincides with the launch of its new Gemini-powered 'Ask Maps' feature.
Read more
xAI Loses 2 Co-Founders as It Preps Grok-3 Launch
Half of the founding team at Elon Musk's xAI has departed the company, according to a report from the Mastodon Social ML Timeline, a significant exodus that has sent shockwaves through the AI sector as the startup prepares for the launch of its anticipated Grok-3 model.
Read more
Mistral's Revenue Soars Past $400M in Push for European AI
Mistral’s annualized revenue run rate now exceeds $400 million, a twenty-fold increase from its $20 million figure just one year ago, according to Financial Times AI. This explosive growth is driven by soaring European demand for sovereign AI alternatives to US-based tech giants.
Read more
OpenAI Enhances ChatGPT Research with Dedicated Document Viewer
OpenAI has deployed a major update to ChatGPT, introducing a full-screen document viewer that enables users to navigate complex, AI-generated reports in a dedicated window, according to a post on the Mastodon Social ML Timeline.
Read more
OpenAI Upgrades API with Agent Skills, Terminal Shell in GPT-5.2
Building an AI agent has been like training a long-distance runner with a thirty-second memory, but OpenAI’s newly upgraded Responses API directly confronts this limitation by adding support for persistent agent skills and a complete terminal shell, according to a VentureBeat AI report.
Read more
ByteDance launches Seedance 2.0 AI video model to rival Sora and Kling
While many AI labs pursue proprietary video generation models like Sora, ByteDance has taken a contrasting open-source approach with its new Protenix-v1, a model that replicates AlphaFold3's performance for biomolecular structure prediction, according to a a blog post report.
Read more
Alibaba Launches Qwen-Image-2.0 to Rival Google's Nano Banana
Alibaba has launched a powerful new vision-language model called Qwen-Image-2.0, a direct shot at Google's Nano Banana, according to a report from Mastodon Social ML Timeline. This marks the latest escalation in the global AI race, as tech giants push further into advanced multimodal systems.
Read more
Microsoft Uncovers Single-Prompt LLM Jailbreak; Copilot Hits 15M Paid Users
While most jailbreak techniques require elaborate multi-step attacks, Microsoft researchers have discovered a new class of single-prompt exploits that can systematically dismantle an LLM’s safety guardrails, according to The Register.
Read more
Oracle Launches Generative AI Certification Amid Java Licensing Concerns
While the cloud AI certification wars have been dominated by AWS and Azure, Oracle has just made a play for relevance with a surprising weapon: a completely free Generative AI certification, according to a technical analysis, practically giving it away to win developer mindshare.
Read more
Anthropic's AI Triggers $285 Billion Market Sell-Off
A single markdown file, not a product or a platform, triggered a market sell-off that erased approximately $285 billion in value during a single trading session, according to a report from Hacker News Newest, an event that dramatically repriced the enterprise software sector.
Read more
Alibaba's Qwen-VL launches 7B vision model with 2K resolution and text rendering
Alibaba's Qwen research team has released Qwen-Image-2.0, a 7-billion-parameter vision model that natively generates and edits images at 2K resolution with advanced text rendering, according to a post on the Mastodon Social ML Timeline.
Read more
Alibaba's Qwen-VL Launches 7B Model with 2K Image Generation
While many AI labs chase scale with ever-larger models, Alibaba's Qwen research team is betting on efficiency, releasing a compact 7-billion-parameter model capable of generating and editing high-fidelity 2K resolution images, according to a report from Mastodon Social ML Timeline.
Read more
Hugging Face Teases Anthropic Collaboration, Ignites Open-Source Debate
While Hugging Face has long been the bastion of open-source AI development, its recent cryptic tease of a collaboration with closed-source giant Anthropic has ignited a fierce debate about the future of AI's foundational principles, according to a post on Mastodon Social ML Timeline.
Read more
Autodesk Sues Google Over "Flow" AI Videomaker Name
For over 25 years, Autodesk has held the trademark for its "Flow" design software. Now, the company is suing Google for using the identical name for its new AI videomaker, alleging trademark infringement, according to a report from Fosstodon AI Timeline.
Read more
Google gains unconditional EU approval for $32B Wiz acquisition
In a regulatory environment where Big Tech acquisitions routinely face stringent conditions, Google secured a remarkably clear path forward, gaining unconditional EU antitrust approval for its landmark $32 billion acquisition of cybersecurity firm Wiz, according to TechMeme.
Read more
Discord assumes all users are underage, requires proof of age
Discord will now assume every user is a minor, requiring them to prove they are an adult through an ID or facial scan, The Register reports. The new policy, set to roll out soon, will use an AI model to potentially infer a user's age, though the burden of proof ultimately falls on you.
Read more
ChatGPT Launches Ad Platform, Joining Select AI Services
While many AI services remain solely focused on subscription revenue, ChatGPT has taken a decisive step toward a hybrid model, launching an advertising platform that will show marketing messages to its US users, according to a report from The Register.
Read more
Showing 20 of 232 articles
Frequently Asked Questions
SectorHQ publishes AI news multiple times daily, with our automated intelligence system monitoring thousands of sources 24/7. Breaking news is published within hours of major announcements, while in-depth analysis follows within the same day.