AI News
The latest developments, breakthroughs, and analysis from the world of artificial intelligence.
ShengShu Launches Vidu Q3, Lands $293M From Alibaba Cloud
China's ShengShu launched Vidu Q3 globally today and closed a $293M Series B led by Alibaba Cloud — putting real pressure on Sora and Veo.
Stanford 2026 AI Index: China Has Erased the US Lead
Stanford HAI's annual AI Index report reveals China and the US are now neck-and-neck on AI performance benchmarks — a historic shift with major implications for business and policy.
DeepSeek R2 Hits 92.7% on AIME With Just 32B Parameters
China's DeepSeek releases an open-weight reasoning model that runs on a single consumer GPU and undercuts Western API pricing by 70%.
GLM-5.1 Tops SWE-Bench Pro, Beating GPT-5.4 and Claude
Z.ai's open-source 754B model achieves state-of-the-art coding benchmarks and runs autonomously for 8 hours straight.
India Bets Big on Sovereign AI With Sarvam's $350M Raise
Sarvam AI is closing a $350M round at a $1.5B valuation as India races to build homegrown AI that works in 22 languages.
OpenAI's $122B Round: The Biggest AI Deal Ever, Decoded
OpenAI closed a record $122 billion fundraise at an $852B valuation — but how much is real cash? Here's what the numbers actually mean.
AI Venture Funding Hits $300B in Q1 2026, Shattering Every Record
Investors poured $300 billion into startups in Q1 2026, with AI claiming 80% of the total — more than all of 2025 combined.
Google Bakes NotebookLM Into Gemini for All Users
Google's AI research tool NotebookLM now lives inside Gemini, bringing persistent memory, source syncing, and project notebooks to every subscriber.
This AI Breakthrough Cuts Energy Use by 100x
Tufts University researchers combine neural networks with symbolic reasoning to slash AI energy consumption by 100x while tripling accuracy.
80,000 Tech Jobs Cut in Q1 2026 — Half Blamed on AI
Tech layoffs hit 80,000 in Q1 2026, with nearly half attributed to AI automation. But experts say the real story is more complicated.
LM Studio Acquires Locally AI, Bringing Local Models to Mobile
LM Studio's acquisition of Locally AI extends on-device AI from desktop to iPhone, iPad, and Mac with cross-device sync.
NVIDIA Bets Big on Physical AI for National Robotics Week
NVIDIA unveiled new Cosmos world models and GR00T robot foundations during National Robotics Week, signaling the next frontier of AI beyond chatbots.
OpenAI's GPT-5.5 'Spud' Is Coming: What We Know
OpenAI's next frontier model, codenamed Spud, finished pretraining in late March. Here's what GPT-5.5 means for developers, businesses, and the AI race.
Revolut Launches AIR: An AI Co-Pilot for Your Money
Revolut's new AI assistant AIR rolls out to 13 million UK users, handling spending insights, card management, and trip planning.
Shopify Launches AI Toolkit: Agents Can Now Run Your Store
Shopify's free, open-source AI Toolkit lets Claude Code, Cursor, and Codex manage stores through natural language.
CoreWeave and Meta Expand AI Cloud Deal to $35 Billion
Meta commits another $21 billion to CoreWeave for AI infrastructure through 2032, featuring Nvidia's Vera Rubin chips and a major shift toward inference workloads.
Google's Gemma 4 Outperforms Models 20x Its Size
Google DeepMind releases Gemma 4 under Apache 2.0, delivering frontier-level reasoning and agentic capabilities in models small enough to run on a laptop.
HeyGen Avatar V: One 15-Second Clip, Unlimited AI Videos
HeyGen's new Avatar V captures your likeness in 15 seconds and generates studio-quality videos in 175+ languages — with full-body movement and consistent identity across every clip.
Perplexity Launches $1B Build Competition as Revenue Surges 50%
Perplexity's pivot to AI agents pushed ARR past $450M in a single month. Now it's daring entrepreneurs to build a billion-dollar company using its Computer platform.
Visa Enables AI Agents to Spend Money Autonomously
Nevermined's integration of Visa Intelligent Commerce and Coinbase's x402 protocol lets AI agents make purchases within user-defined guardrails — no human click required.
Google Launches LiteRT-LM to Run LLMs on Any Edge Device
Google's new open-source framework brings LLM inference to phones, watches, and IoT devices with sub-second response times and full offline support.
Google's TurboQuant Cuts LLM Memory Use by 6x
New compression algorithm shrinks the KV cache to 3 bits per value with zero accuracy loss, making AI inference dramatically cheaper.
Meta Unveils Muse Spark, Its Most Powerful AI Model Yet
Meta Superintelligence Labs debuts Muse Spark, a multimodal AI model built to rival OpenAI and Google across reasoning, health, and coding tasks.
OpenAI, Anthropic, Google Unite to Fight AI Model Copying
Three AI rivals share threat intelligence through the Frontier Model Forum to combat Chinese distillation attacks that siphon billions in R&D.