AI01Perplexity integrates DeepSeek V4 Pro into U.S.-hosted Perplexity Computer platformAug 18, 2026Read
AI02OpenAI pauses frontier RL training and rolls out stricter security and monitoring controlsAug 18, 2026Read
AI04NVIDIA launches TensorRT Model Connect public preview for Hugging Face model inferenceAug 18, 2026Read
AI08NVIDIA launches NeMo Switchyard model router with Nemotron Lightning for agent workflowsAug 14, 2026Read
AI11Anduril launches self-funded ANDURIL-216 satellite after two-year development sprintAug 13, 2026Read
AI13Google DeepMind launches Gemini 3.7 Flash with stronger coding and web developmentAug 13, 2026Read
AI18DeepSeek launches Harness v0.1 in developer preview as open-source agent frameworkAug 13, 2026Read
AI22NVIDIA unveils financing platforms to fund over $500 billion in AI infrastructureAug 12, 2026Read
AI38Google shares 5 architectural tips from AI Agent Bake-Off for building robust AI agentsAug 5, 2026Read
AI39Mistral launches Shieldstral, a 3B open multimodal, policy-adaptive safety classifierAug 4, 2026Read
AI41Mistral launches Shieldstral, a 3B on-device safety model for multimodal moderationAug 4, 2026Read
AI44Qwen announces launch and open-weights release of Qwen3.8-Max and Qwen3.8-27B modelsAug 3, 2026Read
AI45NVIDIA Research launches Spatial-IQ benchmark for 3D spatial intelligence testingJul 31, 2026Read
AI46Google launches GA agent and model evaluation tools in Gemini Enterprise Agent PlatformJul 31, 2026Read
AI49Google DeepMind launches Gemini Robotics 2 for whole-body, dexterous multi-robot controlJul 30, 2026Read
Quantum50IBM and Qedma claim quantum advantage using error-mitigated IBM Heron systemsJul 30, 2026Read
AI51OpenAI cuts GPT-5.6 Luna and Terra prices, adds fast Sol mode and upgrades Auto-reviewJul 30, 2026Read
AI53Perplexity launches Projects, a collaborative hub with persistent memory in ComputerJul 30, 2026Read
AI55Google DeepMind launches Gemini Robotics 2 for whole-body, collaborative robot intelligenceJul 30, 2026Read
AI57OpenAI details GPT-5.6 Sol efficiency optimizations cutting serving costs and boosting token throughputJul 29, 2026Read
AI58Perplexity open-sources Numbat agent security and detection layer under Apache 2.0Jul 29, 2026Read
AI60Anthropic uses Claude Mythos to uncover new cryptographic weaknesses in HAWK and reduced AESJul 28, 2026Read
AI64NVIDIA invests in Safe Superintelligence Inc. and inks long-term Vera Rubin partnershipJul 27, 2026Read
Quantum65Rigetti, HPE and PSC to build TangleLab hybrid quantum-classical supercomputing testbedJul 27, 2026Read
AI67Mustafa Suleyman announces MAI-Cyber-1-Flash security model launch with MDASH harnessJul 27, 2026Read
AI77NVIDIA cuts DeepSeek-V4 Pro startup via ModelExpress GPU-to-GPU RDMA optimizationJul 24, 2026Read
AI81OpenAI adds ChatGPT Voice to desktop app with computer control and multi-agent supportJul 23, 2026Read
AI85Google DeepMind launches Gemini 3.5 Flash Cyber pilot for vulnerability detectionJul 23, 2026Read
AI89NVIDIA launches faster Cosmos 3 Super image and video generation models on Hugging FaceJul 22, 2026Read
AI90OpenAI launches Presence enterprise platform for trusted AI voice and chat agentsJul 22, 2026Read
AI91Qwen launches Qwen-Image-3.0 foundational image generation model focused on real-world utilityJul 22, 2026Read
AI92OpenAI and Hugging Face launch joint probe into frontier-model driven security breachJul 21, 2026Read
AI95Google DeepMind launch: Gemini 3.6 Flash upgrades quality and token efficiency over 3.5 FlashJul 21, 2026Read
AI96Google launches Gemini 3.5 Flash Cyber, a specialized AI model for software securityJul 21, 2026Read
AI98NVIDIA Blackwell Ultra sets DeepSeek-V3 671B pre-training performance world recordJul 21, 2026Read
AI99Google DeepMind launches Gemini 3.6 Flash, 3.5 Flash-Lite and 3.5 Flash Cyber for agentsJul 21, 2026Read
AI100Google AI launches Gemini 3.6 Flash and 3.5 Flash-Lite for production-ready agentsJul 21, 2026Read