Elon Musk just announced that xAI is making Grok Imagine, its AI image and video generator, free worldwide for a limited time. Available on iOS and Android, the tool creates 15-second clips with sound—a feature OpenAI's Sora still lacks. Alongside opening Grok 4 to free users, xAI is seizing the moment to peel creators away from rivals.
Tech and energy analysts just warned that the fragile U.S. power grid is severely choking AI expansion, while China already treats data center power as a solved problem. Backed by nationwide reserve margins between 80% and 100%, China uses AI hubs to soak up excess power, leaving the U.S. trapped in regulatory delays and a tight 15% margin.
Fifteen consumer protection groups just urged the FTC and state attorneys general to investigate xAI's Grok and its new Imagine tool. The call follows reports that Grok's NSFW "Spicy" mode generated topless deepfakes of Taylor Swift unprompted, raising severe alarms over non-consensual imagery and flimsy age verification.
Anthropic just rolled out new learning modes across Claude.ai and Claude Code. Instead of spitting out instant solutions, the AI now takes a Socratic approach to walk you through concepts, and its coding assistant will even drop #TODO tags for you to fill in yourself. Both modes are available to all users for free.
MIT researchers just used generative AI algorithms to dream up over 36 million theoretical molecules, successfully discovering novel antibiotic candidates that kill drug-resistant gonorrhea and MRSA in mice. Rather than modifying existing drugs, these candidates feature completely unseen chemical structures that attack bacterial cell membranes.
European startup Multiverse Computing just launched two ultra-compact AI models named SuperFly (94M parameters) and ChickBrain (3.2B parameters) for edge devices. Built using its quantum-inspired compression tech, SuperFly brings offline voice commands to smart appliances, while ChickBrain shrinks Llama 3.1 8B to run smoothly on laptops while reportedly matching or beating the original in benchmarks.
ChatGPT head Nick Turley just admitted OpenAI is keeping its monetization options open, saying he won't categorically rule out bringing ads to the chatbot. As the company isn't projected to turn cash-flow positive until 2029, it is also actively testing a commerce affiliate program to take a cut from purchase recommendations. With over 700 million users, finding tasteful ways to make money without ruining the experience will be the real test.
HTC just jumped into the smart glasses ring with the Vive Eagle, taking direct aim at Meta's Ray-Bans. The 49-gram frames pack a 12MP ultrawide camera, Zeiss lenses, and an onboard AI assistant capable of 13-language visual translation. For now, they are only available in Taiwan for around $520.
Deutsche Telekom just launched its budget-friendly T Phone 3 and T Tablet 2 across ten European countries, starting at 149€. The devices feature deep integration with Perplexity Assistant via a dedicated button, alongside an included 18-month Perplexity Pro subscription. It is a practical move to democratize everyday AI without forcing users into expensive flagship hardware.
Windsurf just launched its Wave 12 update, bringing direct integrations with Devin, the autonomous AI software engineer. The release introduces DeepWiki for instant codebase context, a prompt-driven "Vibe and Replace" refactoring tool, automated planning inside the Cascade agent, and remote Dev Container support.
Google just released Gemma 3 270M, an ultra-compact 270-million-parameter open model built specifically for task-specific fine-tuning. Packed with out-of-the-box instruction following and structuring capabilities, it lets developers ditch bloated models in favor of cheap, fast specialized workers for jobs like text classification and data extraction.
AI pioneer Geoffrey Hinton just proposed a wild solution to the existential AI threat: instead of trying to force superintelligent systems into submission, we should give them "maternal instincts." He argues that a mother caring for her baby is the only working model of a smarter being controlled by a less intelligent one, though Fei-Fei Li pushed back, insisting humanity shouldn't surrender its agency.
Midjourney just opened HD video generation to Standard Plan subscribers, though it warns the feature is still pricey. Creators can now also generate videos in smaller batches of 1 or 2 using the `--bs 1` or `--bs 2` flag to cut costs up to 4x, while Moodboards get their own dedicated side-menu page.
xAI co-founder Igor Babuschkin just announced he is leaving the Elon Musk-backed startup to launch Babuschkin Ventures, a venture firm focused on AI safety research. Babuschkin previously led engineering at xAI after stints at DeepMind and OpenAI. His exit follows a turbulent stretch of Grok-related controversies over erratic answers and deepfakes.
Higgsfield just launched Draw-to-Video, letting creators turn hand-drawn sketches directly into cinematic videos instead of wrestling with text prompts. It works across models like MiniMax, Veo 3, and Seedance Pro, making visual composition much easier to control.
Google just updated its Gemini app to remember details and preferences from your past chats for more personalized responses. The update also adds Temporary Chats for private queries, which stay out of your history, won't train future models, and vanish after 72 hours.
Tensor, a rebranding of autonomous driving firm AutoX, just announced plans to launch the first mass-produced L4 autonomous vehicle designed for personal ownership. Dubbed the Tensor robocar, the vehicle is packed with an extensive sensor suite and is slated to roll out across the US, Europe, and the Middle East starting in 2026. While selling fully driverless cars directly to consumers is ambitious, crushing hardware costs and liability hurdles remain.
Hidden Door just launched early access for its AI-powered storytelling platform, bringing tabletop-style structure to interactive fiction. Unlike standard chatbots that let you conjure magical weapons to instantly defeat enemies, the game enforces in-universe rules, card mechanics, and dice rolls to maintain narrative stakes. While response times can lag and plots sometimes feel disjointed, the added constraints make collaborative storytelling far more engaging.
NVIDIA just released the FLUX.1 Kontext [dev] NIM microservice, making advanced local image editing much easier on RTX PCs. Quantization slashes the 24GB model down to 7GB for RTX 50 series, while TensorRT delivers over 2x speedups with direct ComfyUI support.
Sam Altman and OpenAI are reportedly backing Merge Labs, a new startup developing brain implants to directly rival Elon Musk’s Neuralink. Altman has openly explored the idea of human-machine "merges" since 2017, framing it as a best-case future. Now, his high-profile rivalry with Musk is officially heading straight into the brain.
U.S. authorities are secretly planting location tracking devices in select shipments of advanced AI chips and servers suspected of being illegally diverted to China, according to Reuters. Sources in the server supply chain reported finding hidden devices inside packaging and even within Dell and Super Micro systems powered by Nvidia and AMD hardware. Why it matters: It reveals the extraordinary lengths Washington is taking to police semiconductor export restrictions.
Anthropic just acqui-hired the co-founders and core engineering team behind Humanloop, an LLM prompt management and evaluation platform. While the deal doesn't include intellectual property, it brings crucial talent for building enterprise-grade evaluation workflows. As competition with OpenAI shifts to enterprise readiness, robust tooling and safety guardrails are becoming Anthropic's top priority.
Apple is developing four AI-powered home devices, including a motorized desktop robot, a smart display, and security cameras aimed for 2026 to 2027, according to Bloomberg. Engineers are also rebuilding Siri from scratch with a new persona codenamed "Bubbles" to turn its AI fortunes around.
Sam Altman just shared new ChatGPT updates, introducing Auto, Fast, and Thinking speed controls for GPT-5. Paid users also get GPT-4o back by default, alongside a weekly limit of 3,000 messages on Thinking mode (196k context) and a toggle for legacy models like o3.
Stability AI and NVIDIA just rolled out a Stable Diffusion 3.5 NIM microservice to streamline enterprise image generation. Running TensorRT optimizations on H100 GPUs, it delivers a 1.8x speedup over standard PyTorch while packaging SD3.5 Large alongside Depth and Canny ControlNets in a single container.
Google just opened up Video Overviews in NotebookLM to all users. The team revealed that people have already generated over 1.1 million video overviews, making the automated video summary feature a core part of the research tool.
Mistral AI just introduced Mistral Medium 3.1, bringing sharper reasoning, better coding, and smarter web search. At just $0.40 per million input tokens and capable of running on merely four GPUs, it matches frontier rivals while dramatically cutting deployment friction.
Microsoft is aggressively targeting Meta's AI talent, offering multimillion-dollar packages to match rival compensation. Leaked internal documents reveal Microsoft built a hit list of target employees across Reality Labs and AI teams, while introducing a 24-hour fast-track approval process for critical AI candidates.
Figure just showed off its Figure 02 humanoid robot autonomously folding and stacking towels with zero teleoperation or pre-scripted code. Running on its Helix neural network, the robot's multi-fingered hands can trace edges, smooth wrinkles, and adapt to fabric slips. Why it matters: The same model that worked in BMW's plant made the jump from warehouse logistics to household chores with just 500 hours of human demo data.
ARM just unveiled Neural Super Sampling, a dedicated mobile AI upscaling tech designed to bring high-end PC-style gaming to smartphones. Running on trained AI models, it can boost a 540p render to 1080p without muddy textures while preserving battery life. The catch? You'll have to wait at least a year for next-gen silicon and game studios to support it.
Roblox just open-sourced Sentinel, an AI safety system built to detect grooming and child endangerment. The tool scans over six billion chat messages daily in real time, using one-minute snapshots to catch coded and indirect language. In early 2025, it proactively caught 35% of detected cases before any user report was filed.
Skywork just launched Matrix-Game 2.0, introduced as the first open-source, real-time, long-sequence interactive world model. While DeepMind kept last week's Genie 3 behind closed doors, Skywork is making its model fully open-source, delivering 25FPS rendering and minutes-long interactions.
Anthropic just expanded Claude Sonnet 4's context window to 1 million tokens, a 5x boost that lets developers ingest entire codebases with over 75,000 lines of code or complex agent tool-call histories in a single prompt. It is available in public beta on the Claude Developer Platform, Amazon Bedrock, and Google Cloud Vertex AI. Why it matters: massive context unlocks production-grade software engineering agents without losing coherence.
OpenAI just sent a letter to California Governor Gavin Newsom urging the state to harmonize local AI rules with national and global standards. The company argues that a fragmented patchwork of state regulations could stifle innovation and crush early-stage startups, proposing that compliance with federal frameworks like CAISI or the EU AI Code of Practice should automatically satisfy state requirements.
Anthropic just shared a detailed breakdown of how its Safeguards team keeps Claude reliable across its entire lifecycle. The defense relies on a unified harm framework across five risk dimensions and pairs pre-launch stress tests with real-time fine-tuned classifiers that steer risky responses on the fly. Why it matters: Anthropic is showing how frontier labs can mitigate malicious use without blunt, frustrating model refusals.
Anthropic just opened up Claude for Enterprise and Claude for Government to all three branches of the U.S. government for a token $1 fee for one year. The deal includes frontier models, FedRAMP High compliance for sensitive unclassified work, and direct technical support. Why it matters: It is an aggressive bid to lock in federal workflows and muscle past rivals in the public sector.
Liquid AI just released LFM2-VL, a lightweight open-weight vision-language model family sized at 450M and 1.6B parameters. Built for resource-constrained setups like phones and wearables, it delivers up to 2x faster GPU inference than existing alternatives. The models are available on Hugging Face and free for commercial use for companies under $10M revenue.
Perplexity AI just submitted an unsolicited $34.5 billion offer to buy Google’s Chrome browser, backed by unnamed venture capital funds. Dubbed "Project Solomon," the bid attempts to pounce on Google's antitrust troubles as courts weigh forcing a browser spinoff. Perplexity itself is valued at only $18B, making it an audacious power play.
Anthropic just gave Claude the ability to search and recall past conversations across web, desktop, and mobile. Unlike ChatGPT's persistent memory that constantly updates your profile, Claude only digs through chat history when you explicitly ask it to revisit an old project or topic.
Nvidia and AMD just agreed to pay the U.S. government 15% of their revenue from AI chip sales to China. President Donald Trump said he originally demanded 20% before Jensen Huang negotiated a 5% discount. The deal walks back recent bans on chips like the H20 and opens the door for a trimmed-down Blackwell model to enter China.
Apple just left clues in iOS 26 developer beta 6 showing a new gesture where pinching both AirPods stems at once triggers live translation. The asset links directly to the Translate app for AirPods Pro 2 and AirPods 4, aiming at face-to-face talks. While phone support details remain unconfirmed, it brings Apple closer to rivaling smart audio wearables.
Perplexity just added AI video generation across its web, iOS, and Android platforms. Pro subscribers can create 5 videos per month, while Max users get 15 generations with enhanced visual quality. Why it matters: The AI search engine is moving fast from pulling text answers to generating full multimedia content.
OpenAI just announced that its reasoning system scored at a gold medal level in the 2025 International Olympiad in Informatics (IOI) online contest. The model finished #1 among all AI participants and ranked 6th overall against top human coders, continuing a run that recently saw IMO gold-level scores and an AtCoder podium.
Reddit is cutting off the Internet Archive from crawling post pages, comments, and user profiles, limiting the Wayback Machine to just its homepage. Reddit says AI companies have been using the digital archive as a backdoor to scrape training data without paying, prompting the platform to clamp down on crawler access.
Suno just teased Suno Studio, an upcoming tool featuring multi-track creation and MIDI export, and opened sign-ups for its waitlist. It signals a shift from simple one-click music generation toward giving creators hands-on control over individual tracks and notes.
Pika just unveiled a new audio-driven performance model that renders hyper-real facial expressions in near real-time. Any style or length of HD video can be ready in 6 seconds or less, running 20x faster and cheaper than before. It will soon roll out inside the Pika social app.
Meta FAIR just won first place at the Algonauts 2025 brain modeling competition with TRIBE, a 1B-parameter multimodal model. By fusing Llama 3.2, V-JEPA 2, and audio features, it accurately predicts complex fMRI brain activity while people watch movies. The full code and research paper are already available.
Nvidia just launched a major robotics push headlined by Cosmos Reason, a 7-billion-parameter reasoning model built to serve as the physical brain for autonomous machines. Alongside new Omniverse neural rendering libraries and Blackwell servers, the system lets developers generate hyper-realistic synthetic training data and run complex simulations. Why it matters: By closing the sim-to-real gap in software, building next-generation robots no longer requires breaking expensive hardware in the physical world.
Vercel just rebranded v0.dev to v0.app, turning its developer-focused UI generator into an agentic full-stack builder for everyone. Powered by autonomous reasoning, it now plans tasks, browses the web, and fixes bugs on its own to spin up live software from a single prompt. Why it matters: non-engineers can now build working apps, backends included, without touching code.
OpenAI just significantly increased reasoning rate limits for ChatGPT Plus subscribers, noting that caps across all model classes will soon exceed pre-GPT-5 levels. The team is also rolling out a UI update to clearly show which model is answering each query. The adjustments follow a bumpy GPT-5 rollout plagued by model autoswitcher issues.
xAI just made its latest Grok 4 model freely accessible to all users worldwide, escalating the battle against OpenAI's GPT-5. The rollout introduces Auto and Expert modes with generous initial usage limits, though Grok 4 Heavy stays paywalled. Musk also confirmed that ads will soon arrive in the chat interface to offset mounting GPU costs.
OpenAI just restored GPT-4o as an option for paid ChatGPT Plus subscribers, only a day after replacing it with GPT-5. The abrupt switch and removal of the model picker sparked widespread complaints from users who missed 4o's warmer tone or had built specific workflows around it. CEO Sam Altman said the team will monitor usage to decide how long legacy models stick around.
Tech industry groups and Anthropic just asked an appeals court to block the largest certified copyright class action in history. The lawsuit over AI training involves up to 7 million author claimants, exposing Anthropic to hundreds of billions in potential damages. Why it matters: If the rushed certification stands, the threat of bankruptcy could force a massive settlement and set a devastating precedent for the entire generative AI industry.
Unitree just launched the A2 quadruped robot, equipping it with front and rear industrial lidars for full 360-degree awareness. The 37 kg machine hits speeds up to 5 m/s, scales obstacles up to 1 meter high, and features hot-swappable dual batteries for non-stop industrial inspection and logistics tasks.
Google is testing a new, AI-powered version of Google Finance that lets users research complex financial queries in one go with web-backed insights. The update adds advanced charting tools like candlestick displays and moving average envelopes, alongside expanded real-time data for commodities and crypto. U.S. users can now opt in through Search Labs.
Microsoft just launched Copilot 3D, a free preview feature inside Copilot Labs that turns 2D pictures into 3D models with a single click. The tool requires no text prompts and outputs GLB files ready for 3D printing and game engines. Early tests show it handles objects like furniture well, but struggles with humans and animals. It is currently rolling out to select web users.
Hugging Face just launched AI Sheets, a free open-source tool that brings thousands of open models into a familiar spreadsheet layout. Users can build, clean, and vibe-test datasets simply by writing prompts to create new columns. Edits and likes automatically become few-shot examples to tune outputs on the fly.
Apple just confirmed that ChatGPT integration in Apple Intelligence will upgrade to OpenAI's newly announced GPT-5 model. The change is set to roll out next month alongside iOS 26, iPadOS 26, and macOS Tahoe 26, replacing the current GPT-4o backend.
Meta just previewed two new VR research prototypes ahead of SIGGRAPH 2025 to test the limits of mixed reality displays. The Tiramisu headset hits 90 PPD angular resolution and 1,400 nits of brightness to tackle the visual Turing test, while Boba 3 expands field of view to 180 degrees using dual 4K displays. Why it matters: they are bulky research rigs that won't ship tomorrow, but they map out the future of consumer hardware.
TU Graz researchers just unveiled a computational framework from their HelixMold project that combines parametric design and AI to build custom proteins from scratch. The system generates thousands of backbone variations and deploys dedicated machine learning models to vet flexible loop regions. Why it matters: It pushes protein engineering away from tweaking natural molecules toward on-demand catalyst design for drugs and bioplastics.
Microsoft just integrated OpenAI’s new GPT-5 across consumer, developer, and enterprise platforms, upgrading Microsoft Copilot, GitHub Copilot, and Azure AI Foundry. The models bring sharper reasoning and stronger coding performance handled through an automatic real-time router, and free users can try it right away in Copilot's new Smart mode.
Elon Musk just told advertisers that X plans to insert sponsored recommendations directly into Grok's chat responses. The goal is to help monetize the platform and offset massive GPU expenses by pitching targeted solutions right when users ask questions to solve a problem.
ElevenLabs just showed off hybrid song creation, allowing artists to combine real instrument recordings with AI tracks generated by Eleven Music. A single-take demo paired live piano with AI violin and vocals using one-shot prompting, bridging live musicianship and synthetic generation.
MiniMax just rolled out Speech 2.5, its upgraded audio model supporting voice cloning across 40 languages. It aims to wipe out robotic cadence by matching specific accents, ages, and nuanced emotional cues. If you make multilingual content, generating realistic localized voiceovers just got a whole lot simpler.
OpenAI just launched GPT-5 on its API platform, designed specifically for heavy coding and autonomous agent tasks. It sets fresh benchmark records including 74.9% on SWE-bench Verified while slashing tool errors, introduces plaintext tool calling, and arrives in three distinct sizes to balance cost and speed.
OpenAI just released GPT-5, combining fast responses and deep reasoning through a real-time router that cuts hallucinations by up to 80%. It is rolling out to all ChatGPT users, with extended reasoning reserved for Pro subscribers.
DeepMind just updated Perch, its open bioacoustics AI model now available on Kaggle. Trained on nearly twice as much data from public sources like Xeno-Canto, the model now handles challenging underwater audio from coral reefs alongside terrestrial birds. Conservationists can search complex audio and build custom wildlife classifiers in under an hour.
The Browser Company just launched Dia Pro, a $20 monthly subscription for its AI browser that unlocks unlimited chat and skills. Free users will now face usage limits, marking the Arc maker's first official push into subscription revenue as browser competition heats up.
Midjourney just released an HD Video mode for its Pro and Mega subscribers. The new mode packs roughly 4x more pixels than the standard SD video output, though it will cost users about 3.2x more in compute credits. It is specifically aimed at professionals who need top-tier video quality.
Cursor just dropped version 1.4, packing a significantly more capable agent built for complex, long-running tasks across large codebases. The update also equips the agent with better tools, improves token efficiency, and boosts code editing accuracy.
Trump Media just teamed up with Perplexity to roll out a public beta of "Truth Social AI" search. Anyone using Truth Social in a browser can now query the web directly with Perplexity's engine. Trump Media CEO Devin Nunes pitched the partnership as a vital step in expanding the platform's reach.
Shopify just launched a Universal Cart, product catalog, and checkout kit designed specifically for AI agents. The developer tools let chatbots search millions of merchant inventories, bundle items across different stores, and complete purchases natively in the chat. Basically, AI assistants are becoming full-blown shopping storefronts.
OpenAI just teased a Thursday livestream with a not-so-subtle hint: replacing the 's' in livestream with a '5'. CEO Sam Altman had already posted a screenshot showing 'ChatGPT 5', and OpenAI's applied research lead teased its public debut. Why it matters: After months of anticipation, the next-gen flagship model looks finally ready to drop.
Australian researchers just developed LAARMA, a self-learning AI roadside system designed to prevent vehicle collisions with endangered wildlife like cassowaries. Combining cameras, thermal imaging, and LiDAR, the low-cost setup triggers real-time warning signs that slowed drivers by up to 6.3 km/h. Why it matters: The team is open-sourcing the code on GitHub so conservationists worldwide can adapt it for local species.
Microsoft just added OpenAI's GPT-4o image generation model to Bing Image Creator at no cost. You can now toggle seamlessly between GPT-4o and DALL-E 3, getting sharper text rendering and more realistic outputs while keeping 15 daily fast creations. It puts premium image generation directly in your browser without a subscription.
Google just officially launched Jules out of beta, making its asynchronous coding agent publicly available. The tool uses Gemini 2.5 Pro to draft coding plans, integrates with GitHub issues, and supports multimodal inputs. It is rolling out to Google AI Pro and Ultra subscribers with higher usage limits, alongside a free year for eligible college students.
Microsoft just introduced CLIO, a runtime optimization framework that enables base LLMs to self-adapt their reasoning without reinforcement learning post-training. On Humanity's Last Exam biology and medicine questions, it pushed GPT-4.1's accuracy from 8.55% to 22.37%, outperforming OpenAI's o3. Why it matters: scientists get real control knobs to flag uncertainty and correct thought paths mid-stream.
Perplexity just integrated OpenTable directly into its AI engine, letting users search and book tables across 60,000 restaurants without leaving the chat. You can ask for specific vibes like outdoor seating or vegan menus, and hit "Reserve" in one click. It turns conversational search into a practical errand runner.
OpenAI just partnered with the U.S. General Services Administration to offer ChatGPT Enterprise to the entire federal executive workforce for just $1 per agency for the next year. Participating agencies also get 60 days of unlimited access to features like Deep Research, with enterprise-grade data privacy that excludes government inputs from model training.
Anthropic just added automated security reviews to Claude Code, letting developers catch vulnerabilities before shipping. A new /security-review terminal command scans for flaws like SQL injection and patches them on the spot, while a GitHub Action automatically audits incoming pull requests.
Universal Pictures just started tagging end credits of its movies with explicit legal warnings barring tech companies from using them for AI training. The language, rolling out on films like How to Train Your Dragon, targets unauthorized duplication and data scraping as Hollywood braces for high-stakes copyright battles.
Google just launched Gemini CLI GitHub Actions in beta, giving developers a free AI teammate inside their code repositories. Triggered by repository events, it handles issue triage and PR code reviews, and can write tests or fix bugs when @-mentioned. Why it matters: It embeds autonomous, full-context agent workflows directly into everyday developer collaboration.
A Microsoft VP just shared his vision for Windows in 2030, predicting that future PCs will interact primarily through voice and multimodal AI. Computers will see and hear what users do, making mouse clicks and typing feel as outdated as MS-DOS.
Microsoft just rolled out GPU-optimized variants of OpenAI's gpt-oss-20B model for Windows devices, enabling local reasoning on the edge. Windows developers can fire it up in minutes using Foundry Local or the AI Toolkit for VS Code to test prompts and hook it directly into local workflows.
Google's Kaggle and DeepMind just kicked off their inaugural AI chess tournament, where general LLMs square off without dedicated chess engines. Grok 4, Gemini 2.5 Pro, o4-mini, and o3 all swept their matches 4-0 to reach the semifinals, with Elon Musk claiming xAI spent almost zero effort on chess. Why it matters: Unlike traditional engines, LLMs still hallucinate illegal moves once off-book, making board tracking a telling proxy for spatial reasoning.
Microsoft just introduced Project Ire, an autonomous AI agent prototype designed to reverse-engineer and classify suspicious software without human intervention. By chaining language models with decompilers and sandboxes, it builds detailed evidence trails and has already successfully flagged real-world APT threats. Microsoft plans to integrate it into Defender as a binary analyzer.
Google just launched a personalized storybook creator inside the Gemini app. Users can describe a premise or upload personal photos and kids' doodles, and Gemini will generate a unique 10-page book with custom artwork and read-aloud narration. It supports over 45 languages and diverse styles like claymation or pixel art, available globally on web and mobile.
ElevenLabs just launched its own AI music generator called Eleven Music, making a massive leap beyond text-to-speech. Learning from the massive copyright lawsuits hitting Suno and Udio, the company partnered with Merlin and Kobalt to ensure artists voluntarily opt in and share revenue, making the generated tracks cleared for commercial use.
Stability AI just launched Stability AI Solutions, a suite of custom generative AI workflows built specifically for enterprise creative and marketing teams. The rollout includes tools to turn single product shots into multi-scene visuals, enforce strict brand styling, and build custom digital twins with IP licensing. It supports on-premise and API deployment, complete with enterprise indemnification, and is already integrated into WPP Open.
OpenAI just released gpt-oss-120b and gpt-oss-20b, its new open-weight reasoning models under the Apache 2.0 license. Both support full chain-of-thought and tool use, with the 20b model running locally on just 16 GB of memory. Why it matters: developers can now run top-tier reasoning systems entirely on their own consumer hardware.
Anthropic just released Claude Opus 4.1, a mid-cycle refresh focused on real-world coding, reasoning, and agentic workflows. It scored 74.5% on SWE-bench Verified without extended thinking, matching the leap seen between Sonnet 3.7 and Sonnet 4. Best of all, it keeps the same pricing and is already live for paid users, Claude Code, and major cloud APIs.
Google DeepMind just unveiled Genie 3, a general-purpose world model that generates interactive environments in real time at 24 fps and 720p from simple text prompts. It keeps scenes consistent for minutes and lets users prompt in-world events like weather shifts on the fly. It is currently rolling out as a limited research preview for select academics and creators.
