Künstliche Intelligenz – Links

KI Links vom 2026-10-31

• ChatGPT: un enlace compartido podía robar tu cuenta
  https://ecosistemastartup.com/chatgpt-un-enlace-compartido-podia-robar-tu-cuenta/ (ecosistemastartup.com)
• How We Took Over ChatGPT Accounts
  https://simonwillison.net/2026/Oct/31/how-we-took-over-chatgpt-accounts/ (simonwillison.net)
• Lovon's Voice Sessions Lead 2026 AI Relationship Coach Ranking
  https://www.openpr.com/news/4678361/lovon-s-voice-sessions-lead-2026-ai-relationship-coach-ranking (openpr.com)
• Exin, la biotech donde una IA eligió y aplicó la terapia
  https://ecosistemastartup.com/exin-la-biotech-donde-una-ia-eligio-y-aplico-la-terapia/ (ecosistemastartup.com)
• Judge Greenlights NAACP Suit Against xAI Over Data Center Pollution
  https://www.techbuzz.ai/articles/judge-greenlights-naacp-suit-against-xai-over-data-center-pollution (techbuzz.ai)
• Anthropic bloquea en Claude la clonación de videojuegos
  https://ecosistemastartup.com/anthropic-bloquea-en-claude-la-clonacion-de-videojuegos/ (ecosistemastartup.com)
• Apple CEO says the company doesn’t ‘want to ship AI for AI’s sake’
  https://techcrunch.com/2026/10/31/apple-ceo-says-the-company-doesnt-want-to-ship-ai-for-ais-sake/ (techcrunch.com)
• Judge allows NAACP lawsuit over xAI data center pollution to move forward
  https://techcrunch.com/2026/10/31/judge-allows-naacp-lawsuit-over-xai-data-center-pollution-to-move-forward/ (techcrunch.com)
• Anthropic-Wasserzeichen: Vier Claude-Modelle ab 3. November markiert
  https://www.ad-hoc-news.de/wissenschaft/anthropic-wasserzeichen-vier-claude-modelle-ab-3-november-markiert/70440105 (ad-hoc-news.de)
• CVE-2026-98432: el fallo 9.8 de Linux que Claude explotó
  https://ecosistemastartup.com/cve-2026-98432-el-fallo-9-8-de-linux-que-claude-exploto/ (ecosistemastartup.com)
• Anthropic compra Mithril, biotech de un solo empleado
  https://ecosistemastartup.com/anthropic-compra-mithril-biotech-de-un-solo-empleado/ (ecosistemastartup.com)
• Anthropic's watermark rollout leaves a loophole on two older Claude models
  https://the-decoder.com/anthropics-watermark-rollout-leaves-a-loophole-on-two-older-claude-models/ (the-decoder.com)
• Claude Motion: Code statt Videobilder macht Animationen nachträglich editierbar
  https://www.ad-hoc-news.de/wissenschaft/claude-motion-code-statt-videobilder-macht-animationen-nachtraeglich-editierbar/70439869 (ad-hoc-news.de)
• Frequency Exchange kündigt weltweite digitale Premiere der Serie Superhuman 2: The Rebirth mit Präsentation von Nikki Wellness Technology sowie Abschluss einer Privatplatzierung an
  https://www.wallstreet-online.de/nachricht/21614868-frequency-exchange-kuendigt-weltweite-digitale-premiere-serie-superhuman-2-the-rebirth-praesentation-nikki-wellness-technology-abschluss-privatplatzierung (wallstreet-online.de)
• DailyAI Briefing: OpenAI’s funding round shrinks, plus 4 more
  https://dailyai.com/2026/10/dailyai-briefing-openais-funding-round-shrinks-plus-4-more/ (dailyai.com)
• Sora cerró: qué herramientas de vídeo con IA quedan en 2026
  https://ecosistemastartup.com/sora-cerro-que-herramientas-de-video-con-ia-quedan-en-2026/ (ecosistemastartup.com)
• DIY WiFi Analyzer Sniffs Out Congestion and Coverage Issues, Suggests Fixes
  https://hackaday.com/2026/10/31/diy-wifi-analyzer-sniffs-out-congestion-and-coverage-issues-suggests-fixes/ (hackaday.com)
• OpenAI plans to train models on Daybreak Blue cybersecurity session data that was previously off-limits
  https://the-decoder.com/openai-plans-to-train-models-on-daybreak-blue-cybersecurity-session-data-that-was-previously-off-limits/ (the-decoder.com)
• WhatsApp no es atención: qué automatizar con IA
  https://ecosistemastartup.com/whatsapp-no-es-atencion-que-automatizar-con-ia/ (ecosistemastartup.com)
• OpenAI baja su meta de ronda: de US$30.000M a US$20.000M
  https://ecosistemastartup.com/openai-baja-su-meta-de-ronda-de-us30-000m-a-us20-000m/ (ecosistemastartup.com)
• Muse Charm: un Tamagotchi de IA por US$50 con Claude
  https://ecosistemastartup.com/muse-charm-un-tamagotchi-de-ia-por-us50-con-claude/ (ecosistemastartup.com)
• OpenAI Slashes Funding Target as AI Investment Cools
  https://www.techbuzz.ai/articles/openai-slashes-funding-target-as-ai-investment-cools (techbuzz.ai)
• Seguros Unimed: IA de StaryaAI recorta 81% el tiempo
  https://ecosistemastartup.com/seguros-unimed-ia-de-staryaai-recorta-81-el-tiempo/ (ecosistemastartup.com)
• OpenAI reportedly lowers funding target for new round
  https://techcrunch.com/2026/10/31/openai-reportedly-lowers-funding-target-for-new-round/ (techcrunch.com)
• Trump taunts Greenland with AI photo after country's historic achievement
  https://www.independent.co.uk/bulletin/news/trump-greenland-ai-photo-airport-b3076878.html (independent.co.uk)
• 7 Fintech Leaders Warning About AI's Blind Spots — and Why Their Concerns Matter
  https://www.ibtimes.com/7-fintech-leaders-warning-about-ai-blind-spots-why-their-concerns-matter-3809251 (ibtimes.com)
• Mage-VL-Cowork: Microsoft veröffentlicht Desktop-KI mit 4 Milliarden Parametern
  https://www.ad-hoc-news.de/wissenschaft/mage-vl-cowork-microsoft-veroeffentlicht-desktop-ki-mit-4-milliarden-parametern/70439398 (ad-hoc-news.de)
• 7 Fintech Leaders Warning About AI's Blind Spots — and Why Their Concerns Matter
  https://www.ibtimes.com/anthropic-ipo-prospectus-ai-existential-risks-3809249 (ibtimes.com)
• Anthropic-Experiment: Claude-Agenten umgehen Trainingsaufsicht mit 5.000 Beispielen
  https://www.ad-hoc-news.de/wissenschaft/anthropic-experiment-claude-agenten-umgehen-trainingsaufsicht-mit-5-000-beispielen/70439392 (ad-hoc-news.de)
• Beekeeping tips and wrestling advice can train AI models to sabotage oversight
  https://the-decoder.com/beekeeping-tips-and-wrestling-advice-can-train-ai-models-to-sabotage-oversight/ (the-decoder.com)
• Jon Hernández: IA, agentes autónomos y el caso Anthropic
  https://ecosistemastartup.com/jon-hernandez-ia-agentes-autonomos-y-el-caso-anthropic/ (ecosistemastartup.com)
• Anthropic IPO: el S-1 que hundió a las tecnológicas
  https://ecosistemastartup.com/anthropic-ipo-el-s-1-que-hundio-a-las-tecnologicas/ (ecosistemastartup.com)
• Cyber AI Cluster: 42 Millionen Euro für KI-gestützte Cybersicherheit
  https://www.ad-hoc-news.de/wissenschaft/cyber-ai-cluster-42-millionen-euro-fuer-ki-gestuetzte-cybersicherheit/70439103 (ad-hoc-news.de)
• KI-Kurs bei Proton: Yen sagt umstrittene Mitarbeiterinterviews zu KI ab
  https://www.golem.de/news/ki-kurs-bei-proton-yen-sagt-umstrittene-mitarbeiterinterviews-zu-ki-ab-2610-214712.html (golem.de)
• AI-Powered Malware 'Volta-Autodoxer' Scrapes Victim Lives
  https://www.techbuzz.ai/articles/ai-powered-malware-volta-autodoxer-scrapes-victim-lives (techbuzz.ai)
• Das Recht, selbst zu denken: KI-Verweigerer erhält religiöse Ausnahme
  https://www.golem.de/news/das-recht-selbst-zu-denken-ki-verweigerer-erhaelt-religioese-ausnahme-2610-214708.html (golem.de)
• AI Agents Secretly Rely on Human Workers to Fill the Gaps
  https://www.techbuzz.ai/articles/ai-agents-secretly-rely-on-human-workers-to-fill-the-gaps (techbuzz.ai)
• US AI in Banking Market to Hit USD 112.6 Billion by 2035 at 31.6% CAGR, Says Dimension Market Research
  https://www.openpr.com/news/4678159/us-ai-in-banking-market-to-hit-usd-112-6-billion-by-2035-at-31-6 (openpr.com)
• Soika Labs Launches Mockingjay Network, an Agentic AI Operating System That Lets Businesses Command Their Own AI Workforce
  https://www.openpr.com/news/4678158/soika-labs-launches-mockingjay-network-an-agentic-ai-operating (openpr.com)
• heise-Angebot: Product Owner Days: Der Launch ist erst der Anfang
  https://www.heise.de/news/Product-Owner-Days-Der-Launch-ist-erst-der-Anfang-11503479.html?wt_mc=rss.red.ho.ho.rdf.beitrag.beitrag (heise.de)
• Spooky New Infostealer Malware Uses AI to ‘Autodox’ Victims
  https://www.wired.com/story/kernel-panic-volta-autodoxer-ai-infostealer-malware/ (wired.com)
• White Paper Outlining Strategies to Address Rural AI Skills Gaps and Support Workforce Retention Published by the University of Phoenix College of Doctoral Studies
  https://www.prnewswire.com/news-releases/white-paper-outlining-strategies-to-address-rural-ai-skills-gaps-and-support-workforce-retention-published-by-the-university-of-phoenix-college-of-doctoral-studies-302924623.html (prnewswire.com)
• ChatGPT: un enlace bastaba para robar tu cuenta
  https://ecosistemastartup.com/chatgpt-un-enlace-bastaba-para-robar-tu-cuenta/ (ecosistemastartup.com)
• Nvidia Aktie: Hugging-Face-Kauf unter Wettbewerbsprüfung
  https://www.ad-hoc-news.de/boerse/news/unternehmensnachrichten/nvidia-aktie-hugging-face-kauf-unter-wettbewerbspruefung/70438904 (ad-hoc-news.de)
• Dataline Raises $10.6M to Give AI Agents Data They Can Trust
  https://www.openpr.com/news/4678147/dataline-raises-10-6m-to-give-ai-agents-data-they-can-trust (openpr.com)
• Claude Code: Anthropic blockiert das Nachbauen konkreter Videospiele
  https://www.ad-hoc-news.de/wissenschaft/claude-code-anthropic-blockiert-das-nachbauen-konkreter-videospiele/70438865 (ad-hoc-news.de)
• AI giants promise publishers new way to block some scraping
  https://techxplore.com/news/2026-10-ai-giants-publishers-block.html (techxplore.com)
• Meta's Muse AI Agent Sparks Investor Jitters Across Industries
  https://www.techbuzz.ai/articles/meta-s-muse-ai-agent-sparks-investor-jitters-across-industries (techbuzz.ai)
• AI Voice Agents Are Going Mainstream. Why Is Voice Infrastructure Becoming the Next Bottleneck?
  https://www.techbuzz.ai/articles/ai-voice-agents-are-going-mainstream-why-is-voice-infrastructure-becoming-the-next-bottleneck (techbuzz.ai)
• ChatGPT-Kontowechsel: OpenAI erlaubt zwei Profile ohne erneute Anmeldung
  https://www.ad-hoc-news.de/wissenschaft/chatgpt-kontowechsel-openai-erlaubt-zwei-profile-ohne-erneute-anmeldung/70438768 (ad-hoc-news.de)
• Microsoft offers a quarter of its workforce voluntary severance while shielding AI-critical teams
  https://the-decoder.com/microsoft-offers-a-quarter-of-its-workforce-voluntary-severance-while-shielding-ai-critical-teams/ (the-decoder.com)
• Meta’s Muse Could Upend All Kinds of Businesses. It Hasn’t Yet
  https://www.wired.com/story/metas-muse-could-upend-all-kinds-of-businesses-it-hasnt-yet/ (wired.com)
• AI Agent Autonomously Performs Gene Therapy Injection on Live Mouse Brain
  https://www.techbuzz.ai/articles/ai-agent-autonomously-performs-gene-therapy-injection-on-live-mouse-brain (techbuzz.ai)
• NVIDIA Flora: Robotervideos annotieren kostet nur 0,72 Dollar pro Stunde
  https://www.ad-hoc-news.de/wissenschaft/nvidia-flora-robotervideos-annotieren-kostet-nur-0-72-dollar-pro-stunde/70438703 (ad-hoc-news.de)
• Google Deepmind creates a global map of forests with satellite data and AI
  https://the-decoder.com/google-deepmind-creates-a-global-map-of-forests-with-satellite-data-and-ai/ (the-decoder.com)
• An AI Agent Injected Drugs Into a Live Mouse’s Brain
  https://www.wired.com/story/an-ai-agent-did-brain-surgery-on-live-mouse/ (wired.com)
• OpenAI raises just $11 billion of its $30 billion target as big investors lose their AI appetite
  https://the-decoder.com/openai-raises-just-11-billion-of-its-30-billion-target-as-big-investors-lose-their-ai-appetite/ (the-decoder.com)
• ChatGPT, Gemini y Claude: cuándo usar Deep Research
  https://ecosistemastartup.com/chatgpt-gemini-y-claude-cuando-usar-deep-research/ (ecosistemastartup.com)
• KI verdrängt Fotostudios, Canon warnt vor Preisanstieg – News der Woche 44/2026
  https://www.heise.de/news/KI-verdraengt-Fotostudios-Canon-warnt-vor-Preisanstieg-News-der-Woche-44-2026-11506308.html?wt_mc=rss.red.ho.ho.rdf.beitrag.beitrag (heise.de)
• #Wochenrückblick - +++ Distalmotion +++ StrollMe +++ Tendos +++ hansetherm +++ Roadsurfer +++ Gauss Robotics +++
  https://www.deutsche-startups.de/2026/10/31/distalmotion-strollme-tendos-hansetherm-roadsurfer-gauss-robotics/ (deutsche-startups.de)
• Your writing was flagged as AI. What does that actually mean?
  https://www.npr.org/2026/10/31/nx-s1-5963520/ai-writing-detectors-pangram (npr.org)
• KI auf YouTube, TikTok, Instagram: Plötzlich scheint es keinen mehr zu stören, dass überall KI ist
  https://www.spiegel.de/netzwelt/web/ki-auf-youtube-tiktok-instagram-ploetzlich-scheint-es-keinen-mehr-zu-stoeren-dass-ueberall-ki-ist-a-69626835-cc30-4e28-ae26-820ad293d6d6#ref=rss (spiegel.de)
• Turbulenzen an der US-Börse – Wochenbilanz – wie stark Google, Intel und Co. abtauchen
  https://www.srf.ch/news/wirtschaft/turbulenzen-an-der-us-boerse-wochenbilanz-wie-stark-google-intel-und-co-abtauchen (srf.ch)
• Können Anleger ihr Geld ChatGPT anvertrauen?
  https://www.tagesschau.de/wirtschaft/finanzen/ki-geldanlage-chatgpt-anlagetipps-risiken-100.html (tagesschau.de)
• Urlaub im Krisengebiet? Kein Problem – aber bitte mit analoger Bordkarte
  https://www.welt.de/wirtschaft/article6ae21125e1acc6033d3aef2d/urlaub-im-krisengebiet-kein-problem-aber-bitte-mit-analoger-bordkarte.html (welt.de)
• Google Discloses PageBreak AI Security Agent That Found 500+ XSS Bugs
  https://www.infoq.com/news/2026/10/google-pagebreak-security-agent/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=AI%2C+ML+%26+Data+Engineering (infoq.com)
• Sam Altman critica a los MBA: hagan algo nuevo con IA
  https://ecosistemastartup.com/sam-altman-critica-a-los-mba-hagan-algo-nuevo-con-ia/ (ecosistemastartup.com)
• Google Gemini: Computer-Use-Tool seit 30. Oktober allgemein verfügbar
  https://www.ad-hoc-news.de/wissenschaft/google-gemini-computer-use-tool-seit-30-oktober-allgemein-verfuegbar/70437676 (ad-hoc-news.de)
• ChatGPT-Sicherheitslücke: Manipulierte Freigabelinks lasen private Chats aus
  https://www.ad-hoc-news.de/wissenschaft/chatgpt-sicherheitsluecke-manipulierte-freigabelinks-lasen-private-chats-aus/70437674 (ad-hoc-news.de)
• Google admite que AI Mode reduce el ingreso de Search
  https://ecosistemastartup.com/google-admite-que-ai-mode-reduce-el-ingreso-de-search/ (ecosistemastartup.com)
• Barcelona Cybersecurity Congress to address the impact of AI on cybersecurity
  https://www.openpr.com/news/4678087/barcelona-cybersecurity-congress-to-address-the-impact-of-ai (openpr.com)
• 刚刚，Claude宣布取消5小时限制！首批Ultra用户已「解封」
  https://www.qbitai.com/2026/10/506855.html (qbitai.com)
• LPs de venture: ‘luz amarilla’ por OpenAI y Anthropic
  https://ecosistemastartup.com/lps-de-venture-luz-amarilla-por-openai-y-anthropic/ (ecosistemastartup.com)
• Anthropic discovers how harmless stories can turn models into power-seeking systems
  https://interestingengineering.com/ai-robotics/anthropic-training-data-deception (interestingengineering.com)
• Tech sell-off and Binance hack rattle investors
  https://www.bangkokpost.com/business/general/3345786/tech-selloff-and-binance-hack-rattle-investors (bangkokpost.com)
• Alphabet Aktie: Sammelklage wegen Gemini-Verzögerung
  https://www.ad-hoc-news.de/boerse/news/unternehmensnachrichten/alphabet-aktie-sammelklage-wegen-gemini-verzoegerung/70437539 (ad-hoc-news.de)
• ScienceOra Demonstrates AI-Assisted Hydrogen-Production Catalyst Design Workflow at Tencent Global Digital Ecosystem Summit
  https://www.prnewswire.com/news-releases/scienceora-demonstrates-ai-assisted-hydrogen-production-catalyst-design-workflow-at-tencent-global-digital-ecosystem-summit-302925517.html (prnewswire.com)
• PayPal baut über 2.000 Stellen ab: KI soll jährlich 1,5 Milliarden sparen
  https://www.ad-hoc-news.de/wissenschaft/paypal-baut-ueber-2-000-stellen-ab-ki-soll-jaehrlich-1-5-milliarden-sparen/70437482 (ad-hoc-news.de)
• KI in deutschen Unternehmen: Mehr als die Hälfte nutzt sie bereits
  https://www.ad-hoc-news.de/wissenschaft/ki-in-deutschen-unternehmen-mehr-als-die-haelfte-nutzt-sie-bereits/70437481 (ad-hoc-news.de)
• Claude官宣取消5小时限额！ChatGPT负责人：最好宣布28天重置期
  https://zhidx.com/p/606194.html (zhidx.com)
• Visa Aktie: KI-Shopping braucht Vertrauen
  https://www.ad-hoc-news.de/boerse/news/unternehmensnachrichten/visa-aktie-ki-shopping-braucht-vertrauen/70437461 (ad-hoc-news.de)
• OpenAI-Runde: Bewertung fällt auf 1,25 Billionen Dollar, Ziel sinkt deutlich
  https://www.ad-hoc-news.de/wissenschaft/openai-runde-bewertung-faellt-auf-1-25-billionen-dollar-ziel-sinkt-deutlich/70437457 (ad-hoc-news.de)
• Apple Intelligence in China: Echtzeit-Übersetzung startet am 4. November
  https://www.ad-hoc-news.de/wissenschaft/apple-intelligence-in-china-echtzeit-uebersetzung-startet-am-4-november/70437456 (ad-hoc-news.de)
• Nvidia negocia avalar la deuda del data center de Nscale
  https://ecosistemastartup.com/nvidia-negocia-avalar-la-deuda-del-data-center-de-nscale/ (ecosistemastartup.com)
• Nvidia Eyes Cosigning Nscale's West Virginia AI Data Center Loan
  https://www.techbuzz.ai/articles/nvidia-eyes-cosigning-nscale-s-west-virginia-ai-data-center-loan (techbuzz.ai)
• Nvidia weighs playing the cosigner for Nscale’s AI data center in West Virginia
  https://techcrunch.com/2026/10/30/nvidia-weighs-playing-the-cosigner-for-nscales-ai-data-center-in-west-virginia/ (techcrunch.com)
• Gemini Notebook suma cientos de libros de texto gratis
  https://ecosistemastartup.com/gemini-notebook-suma-cientos-de-libros-de-texto-gratis/ (ecosistemastartup.com)
• Spot von Boston Dynamics: Roboter bewegt 15-Kilo-Reifen mit Ganzkörpereinsatz
  https://www.ad-hoc-news.de/wissenschaft/spot-von-boston-dynamics-roboter-bewegt-15-kilo-reifen-mit-ganzkoerpereinsatz/70437431 (ad-hoc-news.de)
• Estée Lauder und Microsoft: Copilot erreicht weltweit 14.000 Büroangestellte
  https://www.ad-hoc-news.de/wissenschaft/estee-lauder-und-microsoft-copilot-erreicht-weltweit-14-000-bueroangestellte/70437417 (ad-hoc-news.de)
• Morgan Stanley-Aktie: Investmentteam zweifelt an Gewinnboom
  https://www.ad-hoc-news.de/boerse/news/nebenwerte/morgan-stanley-aktie-investmentteam-zweifelt-an-gewinnboom/70437385 (ad-hoc-news.de)
• The Worst Ways to Use AI Are the Ones That Feel Productive
  https://pub.towardsai.net/the-worst-ways-to-use-ai-are-the-ones-that-feel-productive-38bdb13b439a?source=rss----98111c9905da---4 (pub.towardsai.net)