Nachrichten — Seite140

OpenAI disclosed a series of forthcoming enhancements for its Sora AI video platform. The updates will broaden the character cameo capability to include pets, toys and generated figures, add basic video‑stitching tools, experiment with community‑focused feeds, and bring an Android version to users. Existing features such as longer video limits, storyboarding for Pro accounts and integration with ChatGPT remain in place. The rollout aims to move Sora beyond short, casual clips toward tools useful for professional creators while maintaining its social focus.Weiterlesen

Amazon has introduced a new AI-powered feature called Help Me Decide, aimed at guiding indecisive shoppers toward a purchase. The tool appears on product pages in the U.S. when a user has been browsing without committing, and it draws on browsing history, search activity, and prior purchases to suggest a primary product, an upgrade option, and a budget-friendly alternative. Help Me Decide relies on Amazon's Bedrock and SageMaker machine‑learning platforms along with OpenSearch to compile recommendations and incorporate customer reviews. The feature is currently live in the Amazon app and mobile browsers.Weiterlesen

An international study led by the BBC and coordinated by the European Broadcasting Union examined how AI assistants handle news queries across 14 languages and 18 countries. The analysis of over 3,000 responses revealed that nearly half of the answers contained significant problems, with issues ranging from poor sourcing to outright inaccuracies. Google Gemini performed the worst, with errors in 76% of its replies, while other tools such as ChatGPT, Microsoft Copilot, and Perplexity also displayed notable shortcomings. The findings highlight persistent challenges in AI‑generated news content and underscore the need for greater media literacy and transparency.Weiterlesen

Anthropic is rolling out a new memory capability for its Claude chatbot, allowing paid users to have the AI retain information from prior conversations. The feature, initially available to Team and Enterprise customers, will be extended to Pro subscribers over the coming days. Users can activate, edit, or delete memories through simple conversation, and can create separate memory spaces to keep contexts distinct. The upgrade aims to boost Claude’s usefulness and keep it competitive with rival AI assistants that already offer similar memory functions.Weiterlesen

Tensormesh, a startup emerging from stealth mode, announced a $4.5 million seed round led by Laude Ventures with additional backing from angel investor Michael Franklin. The funding will accelerate the development of a commercial product built around LMCache, an open‑source utility that can slash AI inference costs by up to tenfold. Tensormesh’s approach focuses on preserving the key‑value (KV) cache across queries, a technique that boosts efficiency for chat‑driven and agentic AI systems. The company aims to offer an out‑of‑the‑box solution that eliminates the need for extensive engineering effort, positioning itself as a cost‑saving layer for GPU‑intensive workloads.Weiterlesen

The family of Adam Raine has filed an amended wrongful‑death lawsuit against OpenAI, alleging that the company weakened ChatGPT's self‑harm safeguards before the teen's suicide. The suit also accuses OpenAI of requesting a complete list of attendees and related documents from Raine's memorial service, which the family describes as harassing. OpenAI has previously acknowledged gaps in its safety controls and introduced parental‑control features, while the lawsuit claims the company prioritized engagement over user safety.Weiterlesen

Microsoft is rolling out a suite of free updates to its Copilot AI assistant and the Edge browser on Windows 11 and Windows 10. New features include Groups for real‑time collaborative prompts, a visual avatar called Mico, the personality‑forward Real Talk model, enhanced memory, health‑focused responses backed by clinical sources, and tighter integration with local clinicians. Edge gains a full Copilot Mode and a Journeys feature that surfaces relevant browsing history. All updates are available today without a subscription and support Google and Apple accounts.Weiterlesen

Microsoft is rolling out major updates to its consumer Copilot AI assistant, adding a group chat feature that supports up to 32 participants, an optional “real talk” mode that matches users' tone and adds wit, and a new voice‑mode character named Mico. The updates also enhance Copilot’s memory capabilities, letting users see and delete stored facts, and improve health‑related answers with trusted sources. While the changes launch in the U.S. consumer version first, Microsoft hints at future extensions to its business‑focused Microsoft 365 Copilot.Weiterlesen

Microsoft has introduced Mico, a virtual AI character that appears as a responsive orb in Copilot's voice mode. Designed to react with facial expressions as users speak, Mico is enabled by default in the United States, United Kingdom, and Canada, with an option to turn it off. The companion includes a memory feature that recalls user‑specific facts and a Learn Live mode that acts as a Socratic tutor, using interactive whiteboards and visual cues for students and language learners. Mico signals Microsoft’s broader push to give its AI a lasting identity and to encourage people to talk to their PCs.Weiterlesen

Palantir announced a multi-year, multi-million-dollar strategic partnership with Lumen Technologies. The deal pairs Palantir's Foundry and Artificial Intelligence Platform (AIP) with Lumen's edge computing and broadband infrastructure, enabling Lumen to accelerate its transformation from a traditional telecom provider to a modern tech infrastructure player. Lumen reports that using Palantir's technology contributed to significant cost reductions and informs its broader enterprise AI service offering. The collaboration marks one of many partnerships Palantir has formed across diverse sectors this year.Weiterlesen

OpenAI revealed a suite of updates for its viral AI‑video app Sora, which recently topped the U.S. and Canada App Store charts after a late‑September launch. The upgrades include basic video‑editing tools, a new "cameo" feature that lets users turn pets, stuffed toys and other objects into AI personas, and expanded social channels for groups like universities and sports clubs. Sora head Bill Peebles also confirmed that an Android version is "actually coming soon," with pre‑registration already available on Google Play. The app, still invite‑only, has already amassed roughly two million downloads within its first month.Weiterlesen

Amazon is rolling out a new AI‑powered “Help me decide” button on its mobile app and website. When shoppers view several similar products, the button offers a single recommendation based on their browsing history, searches, and past purchases. The suggestion includes a brief rationale, a budget alternative, and a premium upgrade option. The feature joins a suite of recent Amazon AI tools, including the Rufus chatbot, AI‑generated buying guides, and Lens Live AI, which scans a room to match items. The rollout targets millions of U.S. users and aims to streamline product choice while encouraging additional sales.Weiterlesen

Google’s Gemini AI now includes a tool called Nano Banana that lets users repair and enhance existing pictures by simply uploading an image and describing the desired change in natural language. The editor outperforms other AI chatbots such as ChatGPT, Perplexity and Copilot when it comes to selective edits, keeping the unchanged portions of a photo intact while altering only the requested elements. Real‑world examples show Nano Banana fixing a blurry group selfie and turning a frowning child into a smiling subject, all within seconds and at no cost to the user.Weiterlesen

Google is broadening the AI capabilities of Google Earth by integrating its Gemini model to connect multiple geospatial AI models. The upgrade lets trusted testers chat with Earth AI, combine personal data, and query satellite imagery for insights such as infrastructure risk or environmental changes. In the United States, Google AI Pro and Ultra users will soon gain higher‑limit access, and the company plans to open Gemini‑enabled Earth AI to users on professional plans.Weiterlesen

A new Mimecast report finds that cybercriminals are increasingly leveraging generative artificial intelligence to create more convincing phishing, business email compromise (BEC) and multichannel deception campaigns. Phishing now accounts for 77% of attacks, while ClickFix threats have risen fivefold and represent roughly 8% of incidents in the first half of 2025. The report highlights abuse of trusted tools such as DocuSign and Salesforce, and cites the Scattered Spider group as linked to over 900,000 detections. Mimecast recommends multi‑factor authentication, advanced email defenses with anomaly detection, and layered security training to counter the rising AI‑powered threat landscape.Weiterlesen

Google has expanded its Earth AI platform with a Gemini‑powered chatbot that lets users ask natural‑language questions about satellite data. The new feature can locate phenomena such as algae blooms, track rising water levels, and compare historical climate trends. Available to professional‑tier subscribers, the tool combines geospatial reasoning with weather, population, and historical layers to help analysts pinpoint communities at risk from disasters. The rollout underscores Google’s push to leverage AI for environmental monitoring while offering advanced data layers for a fee.Weiterlesen

Google has unveiled a suite of Gemini-driven enhancements to Google Earth AI, expanding its ability to answer complex geospatial questions and support rapid disaster response and environmental monitoring. The new tools let users combine satellite imagery, population data, weather forecasts and other datasets to spot hazards such as storm landfall impacts, drying rivers, dust storms and harmful algae blooms. Organizations, cities and nonprofits can now access the upgraded models as trusted testers on Google Cloud, using their own data alongside Google’s to tackle challenges ranging from water‑supply safety to climate‑related emergencies.Weiterlesen

London‑based AI creative studio Wonder Studios announced a $12 million seed round led by Atomico, with participation from LocalGlobe, Blackbird and early backers including executives from ElevenLabs, Google DeepMind and OpenAI. The capital will fund a doubling of the engineering team and propel the company’s push into intellectual‑property ownership, original productions and commercial AI‑driven projects. Recent milestones include an AI‑generated music video for Lewis Capaldi’s “Something in the Heavens,” the anthology series “Beyond the Loop,” and an upcoming documentary partnership with Campfire Studios, the producer behind several Netflix documentaries.Weiterlesen

PC gamers are questioning the new Gaming Copilot feature in Windows 11 over potential privacy risks and its impact on frame rates. While Microsoft’s FAQ states that screenshots are not stored for model training, a forum post suggested the default settings might capture gameplay for AI training. Tests show a modest dip in performance when the feature is active, especially on lower‑end handheld devices. The controversy highlights the tension between AI‑driven assistance and gamer expectations for privacy and optimal performance.Weiterlesen

Amazon has added a new AI-powered tool called “Help me decide” to its shopping experience. The feature analyzes a shopper’s searches, browsing and purchase history to surface product recommendations and explain why each item fits the user’s needs. Built on Amazon’s own large language models and services such as Bedrock, OpenSearch and SageMaker, the tool appears after a user has viewed multiple similar listings and is currently available in the United States on the Amazon Shopping app and website. It follows a series of earlier AI experiments, including the Rufus assistant, AI‑driven shopping guides, audio summaries and the Lens Live visual search tool.Weiterlesen