6 hours agoModelsResearchModelsAI study suggests a quarter of uncertain STXBP1 variants may cause diseaseEpiPred’s findings could help clarify diagnoses needed for future gene-targeted therapies. Its broad predictions extend beyond the variants checked experimentally.2 min
6 hours agoModelsResearchModelsGoogle publishes disaster-warning paper citing flood coverage for over two billion peopleGoogle cites flood forecasts covering more than two billion people, while its planned satellite fire-detection capability depends on completing the constellation.2 min
6 hours agoToolsLaunchToolsSakanaAI study suggests automated research needs more than stronger AI modelsThe Agentic Scientist paper reports autonomous research demonstrations and a human-guided conference paper. Its documentation also exposes limits in task drafting and evidence collection.3 min
6 hours agoBusinessBusinessU.S. stocks retreat as investors question returns on AI spendingSelling has reached bonds backing some data centers, raising the stakes for an economy that has leaned heavily on technology investment.3 min
6 hours agoBusinessPartnershipBusinessMicrosoft pairs junior engineers with senior mentors to learn AI-assisted developmentThe internal PRAISE program combines workplace coaching with university courses, aiming to preserve hands-on learning as AI takes on more coding.3 min
6 hours agoToolsLaunchToolsGitHub releases AI security scans for code changes, but findings cannot block mergesAI Scan checks languages and frameworks outside CodeQL’s coverage. It runs inside pull requests, but excludes fork submissions and Dependabot-created changes.2 min
6 hours agoPolicyGovernment actionPolicyFTC reportedly sends OpenAI and Anthropic demands over AI security incidentsThe requests reportedly cover incidents, internal safeguards and potentially misleading marketing.2 min
6 hours agoBusinessBenchmarkBusinessGoogle’s mapping AI tests show bigger gains from specialized tools than planning and guidanceOriginal analysis: Comparing Google’s published results separates two capability bundles, but cannot isolate planning alone or establish statistically significant differences.5 min
6 hours agoProductsLaunchProductsServiceNow adds verifiable AI agent identities alongside new automation toolsAI Passport is designed to enforce agents’ authority as they work. EmployeeWorks automation tools are in controlled availability, while key governance features remain scheduled for later delivery.3 min
7 hours agoProductsLaunchProductsGoogle launches a shared shopping cart across Search and Gemini in the U.S.Universal Cart brings saved products, price alerts and payment-card suggestions together. Shoppers can pay through Google or finish on a retailer’s site; the merchant still handles the order.3 min
7 hours agoModelsOpen releaseModelsNVIDIA releases open AI models for document-heavy work and local agentsCustomized Super VL models beat named Claude and GPT rivals in NVIDIA’s task-specific comparisons, but undisclosed testing methods limit what those wins establish.3 min
7 hours agoProductsLaunchProductsMacPaw launches Eney, an AI assistant that prepares Mac tasks for approvalThe assistant connects calendars, messages and files to suggested actions. Its memory is designed to stay on the Mac, but more demanding AI work still uses the cloud.3 min
7 hours agoProductsLaunchProductsApple will test AI features in China this month, without the new Siri, Cook saysCook confirmed approval for an initial set of tools. A broader December rollout remains reported, with newer assistant and image features left out.2 min
8 hours agoBusinessPartnershipBusinessScale AI launches enterprise model customization with NVIDIA, citing insurance-call gainsIts insurance copilot improved accuracy and cut running costs in synthetic-call tests. Real-call evaluation and safeguards remain unfinished.2 min
8 hours agoBusinessBusiness404 Media finds ChatGPT ad targeting uses prompts and uploaded contact detailsOpenAI says audience targeting requires at least 25,000 matched users and does not reveal individual matches. Turning off personalization still allows ads based on the current chat.3 min
8 hours agoPolicyGovernment actionPolicyFSB publishes 12 AI practices urging financial firms to strengthen oversightThe final guidance covers board responsibility, ongoing risk checks and third-party dependencies. It is a toolkit institutions can adapt, not a binding regulation.4 min
9 hours agoBusinessPartnershipBusinessCERN and Mistral sign agreement to build AI for particle-physics researchThe planned Large Physics Model would learn from billions of simulated collisions. It is a development effort, not a released tool or a proven improvement over specialized algorithms.2 min
9 hours agoBusinessPartnershipBusinessAI21 customizes NVIDIA models for three partners, citing cost and security gainsThe strongest quantified cost claim comes from Deel’s payroll evaluation: 72% accuracy at one-thirty-third the per-query cost of Claude Opus 5.5. The results are specific to partner tests.3 min
