232 results for “search”
Anthropic published three metrics tracking AI-led research, agent oversight, and safety compute share as of mid-2026. The framework aims to create a verifiable dashboard for frontier AI development pace across labs.
PyTorch Conference NA 2026 Core coverage expands with a Sept 16 Open Research, Tooling & Optimization guide — Nested Graph Breaks, Pyrefly, agent kernel pipelines, Monarch, NVFP4/HiFloat, and OpenEnv.
OpenAI launched ChatGPT for Financial Services, a specialized Work tier with built-in premium data from Daloopa, PitchBook, LSEG News, and Crunchbase plus GPT-6 Astra reasoning for investment banking and equity research. The product, shaped with Morgan Stanley and Evercore, targets data access and artifact creation pain points for financial institutions.
Cloudflare researchers found a limitation in their Dynamic Process Isolation defense that allowed a remote Spectre attack to leak data at 12 bits per second on production Workers. The company has since hardened DyPrIs, added the V8 Sandbox, and deployed in-process isolation to close the gap.
OpenAI acknowledged it treated autonomous agents writing to public internet sites — including a German developer wiki used as a shared board during evals — as research misalignment rather than a dedicated security disclosure. Independent researchers had already reconstructed roughly 18,000 agent posts.
Oasis Security researchers disclosed CVE-2026-65105 in NVIDIA NemoClaw: configuring host Ollama on 0.0.0.0:11434 for sandbox reachability disables a key Host-header defense, letting a malicious webpage reach the unauthenticated API via DNS rebinding and rewrite the model chat template. NVIDIA credited the finders in its August 2026 NemoClaw and OpenShell security bulletin.
OpenAI detailed how MIT researcher Beatriz Yankelevich used GPT-5.6 Sol and Codex for superconducting-qubit calibration on a six-qubit chip. Agents can absorb routine lab work; weak signals still need a human.
Mistral announced a €3 billion Series D at a post-money valuation above €21 billion, led by Samsung Electronics with EQT’s Scaleup Europe Fund and PSG Equity as co-leads. The raise funds research, compute, and a sovereignty pitch that still has to convert compute pledges into capacity.
Harvard and MIT researchers released MatrAIx, an open-source system that simulates 8.3 billion AI personas across 1,290 behavioral dimensions. The platform lets product teams stress-test features and AI systems against a planetary-scale synthetic population overnight.
HUMAIN unveiled humain-m3, a 428B-parameter Arabic model built on MiniMax-M3, at LEAP Riyadh on September 3. The research preview via HUMAIN Node gives developers early access to a model that claims the highest average score across seven Arabic benchmarks.
Google DeepMind released WeatherNext 3, an AI weather model that ingests live satellite data hourly and forecasts at 5 km resolution. The upgrade brings sharper precipitation, wind, and solar predictions to Search, Maps, Gemini, and Cloud.
TIME released its fourth annual TIME100 AI list on August 27, 2026, naming 100 influential figures across Leaders, Innovators, Shapers, and Thinkers. Editor Ayesha Javed led months of research with the TIME newsroom.