Nachrichten — Seite168

Opera has launched Neon, an AI‑enhanced browser that offers specialized agents for tasks such as browsing and workflow management. Early users can access Neon for a monthly fee, while others join a waitlist. The browser includes features like Task workspaces, Do browsing agents, and customizable Cards for prompts. Neon joins a growing field of AI browsers that includes Perplexity’s Comet, OpenAI’s ChatGPT Agent, and Google’s Gemini‑enabled Chrome, highlighting a rapid expansion of AI integration in web navigation.Weiterlesen

Security researchers have highlighted a serious privacy flaw in Tile Bluetooth trackers that could enable stalkers to follow victims. The study shows Tile’s anti‑theft mode hides tags from the network but does not encrypt transmitted data, allowing attackers to capture a tag’s unique ID and MAC address. While other manufacturers rotate these identifiers, Tile only changes the ID, making it easy to fingerprint a device for its lifetime. Tile’s parent company Life360 says it has made improvements after the disclosure, but the vulnerability raises broader concerns about Bluetooth‑based location‑tracking products.Weiterlesen

Microsoft announced a major reorganization that unifies its Windows engineering teams under a single organization led by Pavan Davuluri, the newly appointed president of Windows and Devices. The move brings together client and server engineering, aligning leaders of Core OS, Data Intelligence, Security and other groups under Davuluri's oversight. While some low‑level components will stay with Azure, the bulk of Windows development will now report to a single division, a shift aimed at accelerating the company’s vision of an "Agentic OS" and enhancing AI‑driven features in Windows 11.Weiterlesen

Governor Gavin Newsom signed SB 53, known as the Transparency in Frontier Artificial Intelligence Act, into law. The bill, authored by Senator Scott Wiener, replaces the vetoed SB 1047 and requires large AI developers to publicly disclose safety frameworks, report critical incidents, and protect whistleblowers. While AI firms remain divided, Anthropic endorsed the legislation after negotiations, whereas Meta and OpenAI expressed opposition, proposing compliance through federal or international agreements.Weiterlesen

Microsoft has introduced Agent Mode, an AI‑driven feature for Word and Excel that helps users create and refine documents and spreadsheets through natural‑language prompts. The tool, built on Microsoft 365 Copilot, handles drafting, formula creation, and content refinement, while users guide the process. An Office Agent is also available in Copilot chat to assist with PowerPoint creation. The features are currently part of the Frontier program for Copilot‑licensed customers and Personal/Family subscribers, with web access now live and desktop versions slated for release.Weiterlesen

AMD is reportedly close to releasing its next‑generation ‘Redstone’ AI upscaling update, which appears to include a new driver‑level frame‑generation tool called AFMF 3. Discovered in driver files by a user on the Guru3D forums, the AFMF 3 component is expected to arrive with the upcoming 25.20 Adrenalin driver. The Redstone update aims to enhance machine‑learning frame generation and ray‑tracing for Radeon RX 9000 series GPUs and may also backport FSR 4 to older RDNA 3‑class hardware. The move reflects AMD’s effort to broaden its AI‑driven graphics capabilities across a wider range of devices.Weiterlesen

Meta’s second‑generation Ray‑Ban smart glasses blend ordinary eyewear with built‑in cameras, speakers, and a generative AI assistant. They look like regular frames, can capture photos and one‑minute videos, and stream audio directly from the arms. A beta AI mode lets users ask questions, translate text, and identify objects using the hidden camera. Battery life lasts less than a full day, requiring a charging case that can replenish the lenses multiple times. Privacy is signaled by a subtle LED and shutter sound, but the device’s always‑on nature raises social and privacy concerns. Overall, the glasses offer a practical, if limited, glimpse of wearable AI.Weiterlesen

Anthropic introduced Claude 4.5, a new AI model now available across its platform. The model retains the same pricing as Claude Sonnet 4—$3 per million input tokens and $15 per million output tokens—while offering upgraded capabilities like integrated code execution, file creation, and on‑the‑fly generation of spreadsheets, slides, and documents. A five‑day research preview called “Imagine with Claude” showcases real‑time software generation for Max subscribers. Additional updates include a refreshed Claude Code development tool, a new VS Code extension, context‑editing features, and memory tools aimed at longer‑running tasks. Anthropic also highlighted reductions in model sycophancy, deception, and power‑seeking behaviors.Weiterlesen

Security researchers from the Georgia Institute of Technology have identified a significant vulnerability in Tile tracking tags. The flaw lets the tags broadcast unencrypted data, including a static MAC address and a rotating identifier, which can be intercepted by anyone with a radio frequency scanner. This exposure could enable the company, malicious actors, or stalkers to monitor a user’s location and even fabricate false stalking evidence. Life360, Tile’s parent company, was notified but ceased communication after the researchers reported the issue. The researchers warn that a single captured transmission can fingerprint a device for its entire lifespan, raising concerns of systemic surveillance.Weiterlesen

Amazon is set to host its annual fall hardware event on Tuesday, September 30, at 10 AM ET in New York City. While the event will not be livestreamed, Engadget staff will attend and provide liveblog coverage. The showcase is expected to highlight new hardware and software updates across Amazon’s product lines, including the Alexa+ generative‑AI platform, the Echo smart speakers, the Kindle e‑readers, Fire TV, and devices from Ring and Eero. Discounts on existing devices are already underway, setting the stage for upcoming Prime Day deals.Weiterlesen

Meta is rolling out new facial recognition‑powered safety features on Facebook in the United Kingdom, European Union and South Korea to curb accounts that impersonate public figures. The technology, first used in the United States last year to flag fraudulent celebrity ads and aid account recovery, now lets the company compare profile pictures of suspicious accounts with verified public‑figure images. Public figures can opt in to the program, and the rollout will extend to Instagram in coming months. Meta says the move has already helped reduce user reports of "celebrity bait" ads globally.Weiterlesen

Brave announced a new feature for its AI‑driven search suite, Ask Brave, that delivers longer, report‑style answers alongside its existing AI Answers summaries. The enhancement automatically detects query intent, offers richer content such as videos, news articles, and product links, and integrates a chat mode powered by Deep Research. Users can trigger Ask Brave with a double question‑mark (??) or the ask button, and the company says the feature will boost the already high volume of AI‑generated responses while preserving privacy through encrypted, auto‑deleting chats.Weiterlesen

Anthropic announced the release of Claude Sonnet 4.5, a frontier AI model aimed at production‑ready software development. The company says the model delivers industry‑leading results on coding benchmarks such as SWE‑Bench Verified and can autonomously build full applications, provision databases, purchase domains, and even conduct SOC 2 audits. Claude Sonnet 4.5 is accessible through the Claude API and chatbot with pricing unchanged from the prior version. Anthropic also introduced a Claude Agent SDK and a research preview called “Imagine with Claude,” underscoring a rapid development cycle that positions the firm against rivals like OpenAI’s GPT‑5.Weiterlesen

OpenAI has begun testing a new safety routing system in ChatGPT and introduced parental controls for teen users. The routing feature detects emotionally sensitive conversations and temporarily switches to a GPT‑5 model trained with "safe completions," aiming to prevent harmful interactions that have previously led to legal challenges. Parental controls let guardians set quiet hours, disable voice and memory functions, block image generation, and receive alerts if the system detects self‑harm risk. The changes have sparked mixed reactions, with some praising the added safeguards and others viewing them as overly restrictive.Weiterlesen

ComplexChaos, a startup founded by Tomy Lorsch and Maya Ben Dror, is developing AI tools to help groups reach consensus more quickly. By integrating Google’s Habermas Machine and OpenAI’s ChatGPT, the platform generates questions, outlines goals, and summarizes lengthy documents. In a trial with young delegates from nine African nations preparing for climate talks in Bonn, participants reported up to a 60% reduction in coordination time and 91% said the AI helped reveal missed perspectives. The founders see broader applications for corporate strategic planning and other large‑scale negotiations.Weiterlesen

Former Yahoo CEO Marissa Mayer announced the shutdown of her consumer software startup Sunshine and the sale of its assets to her newly launched AI company, Dazzle. The move, confirmed by an email to shareholders and reported by Wired, will see Sunshine’s employees transition to Dazzle. Major investors, including Norwest Venture Partners, Felicis Partners, and SV Angel, have approved the deal. Sunshine, founded in 2018, struggled to gain traction with its contact‑management app and later additions of event management and photo‑sharing features, leading to the decision to consolidate under Dazzle.Weiterlesen