Nachrichten — Seite35

OpenAI has rolled out Chronicle, a research‑preview feature for its Codex desktop app on Apple Silicon Macs. The tool periodically snaps screenshots, sends them to OpenAI’s servers for OCR and visual analysis, and saves text summaries as unencrypted Markdown files on the user’s machine. Available only to ChatGPT Pro subscribers paying $100 or more per month, Chronicle is barred from the EU, UK and Switzerland. While the feature promises seamless context for the AI assistant, it also raises fresh concerns about data privacy and regulatory compliance.Weiterlesen

The National Security Agency has begun using Mythos Preview, Anthropic’s advanced AI model that the company kept out of public release. The move follows a clash with the Department of Defense, which had labeled Anthropic a supply‑chain risk after the firm refused unrestricted Pentagon access to its tools. Anthropic limited Mythos to roughly 40 vetted organizations, naming only a handful publicly; the NSA is now among the undisclosed users, employing the model to scan for exploitable cyber vulnerabilities. The development coincides with a tentative thaw in Anthropic’s ties to the Trump administration, highlighted by a recent meeting with senior White House officials.Weiterlesen

Around 10 a.m. ET (3 p.m. GMT) on April 20, 2026, OpenAI’s flagship chatbot ChatGPT and its code‑generation tool Codex experienced a partial outage. Users reported problems ranging from login failures to missing conversation histories and sluggish image generation. A poll conducted by TechRadar showed 63 percent of respondents struggled to access old chats, while 27 percent could not sign in. By early afternoon the number of complaints on Downdetector began to fall, and OpenAI’s status page indicated ongoing investigation but no clear timeline for full restoration.Weiterlesen

Anthropic's new Mythos AI model has sparked concern among security experts after data from CrowdStrike showed AI‑enabled cyber attacks jump 89 percent in 2025. The model's ability to automate vulnerability hunting could overwhelm defenders, with internal warnings that companies may discover more flaws than they can patch. Recent incidents, including a Chinese‑linked AI espionage campaign that used Anthropic's Claude Code to breach dozens of high‑profile targets, underscore the growing threat. Analysts argue that granting AI agents unrestricted access to data, the internet, and external communication creates a “lethal trifecta” for hackers.Weiterlesen

Google’s Gemini AI now lives in most of its consumer services, from Gmail to Search, often surfacing suggestions that feel intrusive. Users can pull back the AI by disabling Google Workspace smart features and turning off experimental labs. The process involves a few clicks inside account settings, but it also means losing conveniences like calendar event auto‑creation and personalized search results. This guide walks readers through the exact steps to mute the AI while explaining what functionality they’ll give up.Weiterlesen

OpenAI announced two back‑to‑back acquisitions: personal‑finance startup Hiro and media company TBPN, a business‑talk show producer. The moves come as the company faces pressure to expand beyond ChatGPT and to counter Anthropic’s gains in the enterprise market. Analysts see the deals as talent grabs aimed at diversifying OpenAI’s product lineup and improving its public image amid growing scrutiny.Weiterlesen

The National Security Agency has begun using Anthropic’s new Mythos Preview, a general‑purpose language model touted for its strength in computer‑security tasks. Sources familiar with the rollout say the NSA is one of roughly 40 agencies granted access and that usage is expanding within the department. The move comes despite a months‑long feud between the AI firm and the Pentagon, a February order from former President Trump to halt government use of Anthropic services, and ongoing lawsuits over the company’s designation as a supply‑chain risk.Weiterlesen

Anthropic announced Claude Design, an AI‑powered design platform that lets users create prototypes, presentations and marketing assets by simply describing what they need. Built on the company’s Claude Opus 4.7 vision model, the tool is now in research preview for paid Claude subscribers and will roll out gradually. It promises to cut design time, automate brand‑system application and enable real‑time collaboration, positioning itself as a bridge between seasoned designers and non‑technical creators.Weiterlesen

A growing cohort of AI researchers, tech leaders and content creators—dubbed “doom influencers”—is pushing warnings about the risks of increasingly powerful artificial intelligence. Their messages, ranging from job displacement to existential threats, are gaining traction as companies like Anthropic hold back the release of its most advanced model, Mythos, limiting access to a handful of vetted partners. Governments in the UK, Canada and India are also taking note, sparking a broader debate on how to balance rapid AI progress with safety and regulation.Weiterlesen

Google announced that its Gemini Notebooks feature, previously limited to paid AI subscribers, is now available to anyone using the free version of Gemini on the web. The move lets users store conversations, files and sources in a dedicated workspace that Gemini can draw on for context. Free accounts can add up to 50 sources per notebook, while paid tiers enjoy higher limits. The feature, which syncs with Google’s NotebookLM research tool, is currently limited to the web platform, with mobile and desktop apps slated for future release.Weiterlesen

Amsterdam‑based startup Schematik, founded by former maker Samuel Beek, has secured $4.6 million in seed funding from Lightspeed Venture Partners. The company’s new AI‑driven platform promises to guide users through the entire hardware‑building process, from design to parts sourcing, and even offers step‑by‑step assembly instructions. Early adopters have already used the tool to create everything from MP3 players to custom Tamagotchi‑style bots. Anthropic recently announced a Bluetooth API that lets developers connect its Claude model to physical devices, a move that aligns closely with Schematik’s vision. The startup aims to lower the barrier to entry for electronics creation while keeping safety at the forefront.Weiterlesen

Tools for Humanity’s World verification platform announced a global rollout of its “proof of human” technology, beginning with Tinder and expanding to concert ticketing, Zoom calls and document signing. The move aims to authenticate real users while preserving anonymity as AI‑generated content floods digital services. World will embed a verified World ID badge on Tinder profiles and partner with Ticketmaster, Eventbrite, 30 Seconds to Mars and Bruno Mars to reserve tickets for verified humans. The company also unveiled new verification tiers, including low‑friction selfie checks, to address scaling challenges.Weiterlesen

Nvidia chief Jensen Huang told listeners on the Dwarkesh Podcast that DeepSeek’s plan to run its upcoming V4 foundation model on Huawei’s Ascend 950PR processor would be “a horrible outcome” for the United States. The Chinese lab’s migration from Nvidia’s CUDA software to Huawei’s CANN framework threatens the hardware‑software dependency that has underpinned America’s AI dominance. Huang’s remarks come as U.S. lawmakers consider adding DeepSeek to the export‑control entity list, and as the industry watches whether Huawei’s chips can close the performance gap with Nvidia’s GPUs.Weiterlesen

Anthropic CEO Dario Amodei met with senior officials from the Trump administration this week, signaling a possible thaw in relations after the Pentagon labeled the AI firm a supply‑chain risk. Treasury Secretary Scott Bessent and White House chief of staff Susie Wiles joined the discussion, which the White House described as productive. While the Pentagon continues to challenge Anthropic’s models for military use, other agencies appear eager to explore the company’s technology for cybersecurity, AI safety and maintaining America’s lead in the AI race.Weiterlesen

Cerebras Systems, the San Diego‑based AI chip maker, filed to go public this week, targeting a mid‑May offering. The filing follows two massive funding rounds that lifted the company’s valuation to $23 billion and brought total capital raised to $2.1 billion. Recent contracts with Amazon Web Services and a reported $10 billion deal with OpenAI underscore the firm’s rapid ascent. Revenues hit $510 million in 2025, while net income reached $237.8 million, according to the prospectus. The IPO marks a new chapter for a startup that once withdrew a 2024 filing amid a federal review.Weiterlesen

Kevin Weil, OpenAI's former chief product officer who recently led the company’s new AI workspace for scientists called Prism, announced his exit on Friday. The departure coincides with OpenAI’s decision to dissolve Prism, folding its roughly 10‑person team into the Codex division. The move is part of a broader effort to streamline product offerings and focus on enterprise and coding tools as the company prepares for an IPO. OpenAI also confirmed the launch of GPT‑Rosalind, a suite of models aimed at accelerating life‑science research.Weiterlesen

Anthropic unveiled Claude Design on Friday, its first proprietary AI design platform aimed at workplace creators. The research‑preview tool lets users generate slide decks, social‑media graphics, and app or web interface mockups with fine‑grained controls for spacing, color and layout. Powered by the new Opus 4.7 model, Claude Design can analyze a company’s codebase and brand assets to ensure visual consistency. The service is now available to Pro, Max, Team and Enterprise subscribers, positioning Anthropic’s AI offerings squarely in the business‑productivity space rather than the consumer‑focused creative market.Weiterlesen

OpenAI unveiled GPT‑Rosalind, its first large‑language model built specifically for life‑science research. Named for DNA pioneer Rosalind Franklin, the system is designed to help scientists sort through massive data sets, generate hypotheses and speed the development of new medicines. OpenAI says the model can cut the 10‑ to 15‑year timeline typical for U.S. drug approval by improving target selection and experiment design. Available now as a research preview through a trusted‑access platform, GPT‑Rosalind also includes safeguards against misuse, as the company faces a copyright lawsuit from Ziff Davis.Weiterlesen

Anthropic introduced Claude Design, a research‑preview app that lets subscribers generate prototypes, slides, and full‑scale designs using its latest vision model, Opus 4.7. The tool starts with a text prompt and lets users refine outputs through conversation, inline comments and custom sliders. It can ingest an organization’s existing design assets to adopt its colors and typography automatically, and supports image uploads, document imports, and web captures. Claude Design is available to Pro, Max, Team and Enterprise customers and competes with recent AI assistants from Adobe and Canva.Weiterlesen

Anthropic’s newest AI model, Claude Mythos Preview, is being pitched as a cyber‑security tool that can spot vulnerabilities in major software platforms. The rollout follows months of acrimonious exchanges with the Trump administration, which branded the company a “radical left, woke” threat to national security. With Apple, Nvidia and JPMorgan Chase already on board, and senior officials reportedly briefed on the model’s capabilities, Anthropic appears to be rebuilding bridges with the Pentagon, the White House and other agencies.Weiterlesen