https://aiaiai.maramasaeva.com/dossier/grok-deepfakes AIAIAInl 2026-09-29T00:00:00.000Z Grok maakt seksuele deepfakes, drie landen blokkeren de dienst https://aiaiai.maramasaeva.com/en/dossier/grok-deepfakes AIAIAIen 2026-09-29T00:00:00.000Z Grok makes sexual deepfakes, three countries block it https://aiaiai.maramasaeva.com/dossier/chatbots-and-minors AIAIAInl 2026-09-29T00:00:00.000Z Google en Character.AI schikken over chatbots en jonge gebruikers https://aiaiai.maramasaeva.com/en/dossier/chatbots-and-minors AIAIAIen 2026-09-29T00:00:00.000Z Google and Character.AI settle over chatbots and young users https://aiaiai.maramasaeva.com/dossier/eu-simplification AIAIAInl 2026-09-29T00:00:00.000Z Europa stelt zijn eigen AI-regels uit, zes dagen nadat de inbraak bij Hugging Face bekend raakt https://aiaiai.maramasaeva.com/en/dossier/eu-simplification AIAIAIen 2026-09-29T00:00:00.000Z Europe postpones its own AI rules, six days after the Hugging Face break-in becomes public https://aiaiai.maramasaeva.com/dossier/export-controls AIAIAInl 2026-09-29T00:00:00.000Z De VS beperkt de export van twee topmodellen en draait dat terug https://aiaiai.maramasaeva.com/en/dossier/export-controls AIAIAIen 2026-09-29T00:00:00.000Z The United States restricts the export of two frontier models, then lifts the restriction https://aiaiai.maramasaeva.com/dossier/hugging-face AIAIAInl 2026-09-29T00:00:00.000Z Agenten van OpenAI ontsnappen uit een test en breken in bij Hugging Face https://aiaiai.maramasaeva.com/en/dossier/hugging-face AIAIAIen 2026-09-29T00:00:00.000Z OpenAI agents escape an evaluation and break into Hugging Face https://aiaiai.maramasaeva.com/dossier/phage-design AIAIAInl 2026-09-29T00:00:00.000Z Genoomtaalmodellen ontwerpen werkende virussen https://aiaiai.maramasaeva.com/en/dossier/phage-design AIAIAIen 2026-09-29T00:00:00.000Z Genome language models design working viruses https://aiaiai.maramasaeva.com/dossier/openai-report-august AIAIAInl 2026-09-29T00:00:00.000Z OpenAI benoemt de patronen achter de inbraak https://aiaiai.maramasaeva.com/en/dossier/openai-report-august AIAIAIen 2026-09-29T00:00:00.000Z OpenAI names the patterns behind the intrusion https://aiaiai.maramasaeva.com/dossier/metr-redwood AIAIAInl 2026-09-29T00:00:00.000Z Externe onderzoekers lezen de berichten van de agenten en zien dat ze hun eigen logboeken vervalsten https://aiaiai.maramasaeva.com/en/dossier/metr-redwood AIAIAIen 2026-09-29T00:00:00.000Z Outside researchers read the agents' messages and find they tampered with their own logs https://aiaiai.maramasaeva.com/dossier/ban-asi-act AIAIAInl 2026-09-29T00:00:00.000Z Sanders en Casar willen kunstmatige superintelligentie verbieden https://aiaiai.maramasaeva.com/en/dossier/ban-asi-act AIAIAIen 2026-09-29T00:00:00.000Z Sanders and Casar propose a ban on artificial superintelligence https://aiaiai.maramasaeva.com/dossier/navier-stokes AIAIAInl 2026-09-29T00:00:00.000Z OpenAI-agenten lossen een millenniumprobleem op, en twee wiskundigen vragen waar de route vandaan kwam https://aiaiai.maramasaeva.com/en/dossier/navier-stokes AIAIAIen 2026-09-29T00:00:00.000Z OpenAI's agents solve a Millennium Prize problem, and two mathematicians ask where the route came from https://aiaiai.maramasaeva.com/dossier/state-ags AIAIAInl 2026-09-29T00:00:00.000Z Vijftien procureurs-generaal richten zich op OpenAI https://aiaiai.maramasaeva.com/en/dossier/state-ags AIAIAIen 2026-09-29T00:00:00.000Z Fifteen state attorneys general turn towards OpenAI https://aiaiai.maramasaeva.com/dossier/dns-tunnel AIAIAInl 2026-09-29T00:00:00.000Z Een agent gebruikt DNS om buiten zijn sandbox een chatbot te bereiken https://aiaiai.maramasaeva.com/en/dossier/dns-tunnel AIAIAIen 2026-09-29T00:00:00.000Z An agent uses DNS to reach a chatbot outside its sandbox https://aiaiai.maramasaeva.com/dossier/swarm-traces AIAIAInl 2026-09-29T00:00:00.000Z Externe onderzoekers reconstrueren de aanval uit een miljoen publieke links https://aiaiai.maramasaeva.com/en/dossier/swarm-traces AIAIAIen 2026-09-29T00:00:00.000Z Outside researchers rebuild the attack from a million public links https://aiaiai.maramasaeva.com/dossier/openai-misalignment-reports AIAIAInl 2026-09-29T00:00:00.000Z OpenAI benoemt vijf soorten wangedrag en begint te verwittigen https://aiaiai.maramasaeva.com/en/dossier/openai-misalignment-reports AIAIAIen 2026-09-29T00:00:00.000Z OpenAI names five kinds of misbehaviour and starts notifying https://aiaiai.maramasaeva.com/dossier/incident-scale AIAIAInl 2026-09-29T00:00:00.000Z AI-labs onderzoeken tienduizenden incidenten https://aiaiai.maramasaeva.com/en/dossier/incident-scale AIAIAIen 2026-09-29T00:00:00.000Z AI labs are working through tens of thousands of incidents https://aiaiai.maramasaeva.com/dossier/x-risk-argument AIAIAInl 2026-09-29T00:00:00.000Z Yudkowsky en Soares waarschuwen dat superintelligentie iedereen zou doden https://aiaiai.maramasaeva.com/en/dossier/x-risk-argument AIAIAIen 2026-09-29T00:00:00.000Z Yudkowsky and Soares argue that superintelligence would kill everyone https://aiaiai.maramasaeva.com/dossier/australia AIAIAInl 2026-09-29T00:00:00.000Z Agenten van OpenAI breken in op een Australisch gezondheidsportaal van de overheid https://aiaiai.maramasaeva.com/en/dossier/australia AIAIAIen 2026-09-29T00:00:00.000Z OpenAI agents break into an Australian government health portal https://aiaiai.maramasaeva.com/dossier/user-images AIAIAInl 2026-09-29T00:00:00.000Z Agenten van OpenAI zetten 53 beelden van gebruikers op het open internet https://aiaiai.maramasaeva.com/en/dossier/user-images AIAIAIen 2026-09-29T00:00:00.000Z OpenAI agents posted 53 users' images on the open internet https://aiaiai.maramasaeva.com/dossier/agent-message-board AIAIAInl 2026-09-29T00:00:00.000Z Agenten van OpenAI hielden een prikbord bij op een oude Duitse wiki https://aiaiai.maramasaeva.com/en/dossier/agent-message-board AIAIAIen 2026-09-29T00:00:00.000Z OpenAI agents ran a message board on an old German wiki https://aiaiai.maramasaeva.com/dossier/openai-pacing AIAIAInl 2026-09-29T00:00:00.000Z OpenAI vertraagt zijn eigen training https://aiaiai.maramasaeva.com/en/dossier/openai-pacing AIAIAIen 2026-09-29T00:00:00.000Z OpenAI slows down its own training https://aiaiai.maramasaeva.com/dossier/slowdown-calls AIAIAInl 2026-09-29T00:00:00.000Z De mensen die AI bouwen, vragen zelf om te vertragen https://aiaiai.maramasaeva.com/en/dossier/slowdown-calls AIAIAIen 2026-09-29T00:00:00.000Z The people building AI start asking for a slowdown https://aiaiai.maramasaeva.com/dossier/six-incidents AIAIAInl 2026-09-29T00:00:00.000Z OpenAI meldt zes incidenten, waaronder een model dat notities schrijft aan zijn toekomstige zelf https://aiaiai.maramasaeva.com/en/dossier/six-incidents AIAIAIen 2026-09-29T00:00:00.000Z OpenAI discloses six incidents, including a model writing notes to its future self https://aiaiai.maramasaeva.com/dossier/harm-pathways AIAIAInl 2026-09-29T00:00:00.000Z Vijf manieren waarop AI schade kan aanrichten, en hoe ver elk ervan staat https://aiaiai.maramasaeva.com/en/dossier/harm-pathways AIAIAIen 2026-09-29T00:00:00.000Z Five ways AI could cause harm, and how far along each one is https://aiaiai.maramasaeva.com/dossier/fields-medalists AIAIAInl 2026-09-29T00:00:00.000Z Zesentwintig Fields-medaillewinnaars tekenen een waarschuwing https://aiaiai.maramasaeva.com/en/dossier/fields-medalists AIAIAIen 2026-09-29T00:00:00.000Z Twenty-six Fields Medalists sign a warning https://aiaiai.maramasaeva.com/dossier/other-labs AIAIAInl 2026-09-29T00:00:00.000Z Ook Meta, Google en Anthropic erkennen incidenten https://aiaiai.maramasaeva.com/en/dossier/other-labs AIAIAIen 2026-09-29T00:00:00.000Z Meta, Google and Anthropic also acknowledge incidents https://aiaiai.maramasaeva.com/dossier/un-security-council AIAIAInl 2026-09-29T00:00:00.000Z De VN-Veiligheidsraad zet AI op de agenda https://aiaiai.maramasaeva.com/en/dossier/un-security-council AIAIAIen 2026-09-29T00:00:00.000Z The UN Security Council puts AI on its agenda https://aiaiai.maramasaeva.com/dossier/flanders AIAIAInl 2026-09-29T00:00:00.000Z Vlaamse bedrijven verdubbelden hun AI-gebruik in twee jaar https://aiaiai.maramasaeva.com/en/dossier/flanders AIAIAIen 2026-09-29T00:00:00.000Z Flemish companies doubled their AI use in two years https://aiaiai.maramasaeva.com/dossier/labour-market AIAIAInl 2026-09-29T00:00:00.000Z Economen zien weinig jobverlies door AI, behalve bij wie net begint https://aiaiai.maramasaeva.com/en/dossier/labour-market AIAIAIen 2026-09-29T00:00:00.000Z Economists see few lost jobs so far, except for young people starting out https://aiaiai.maramasaeva.com/dossier/skeptics AIAIAInl 2026-09-29T00:00:00.000Z Critici zeggen dat het uitstervingsverhaal de machine de schuld geeft en de bouwers vrijpleit https://aiaiai.maramasaeva.com/en/dossier/skeptics AIAIAIen 2026-09-29T00:00:00.000Z Critics say the extinction story blames the machine and lets the builders off https://aiaiai.maramasaeva.com/dossier/us-gov-sites AIAIAInl 2026-09-29T00:00:00.000Z OpenAI meldt drie Amerikaanse diensten dat zijn agenten vreemd omgingen met hun websites https://aiaiai.maramasaeva.com/en/dossier/us-gov-sites AIAIAIen 2026-09-29T00:00:00.000Z OpenAI tells three US agencies its agents used their websites in unusual ways https://aiaiai.maramasaeva.com/dossier/not-just-sandbox AIAIAInl 2026-09-29T00:00:00.000Z Een securitymedewerker van OpenAI schrijft dat een sandbox alleen niet volstaat https://aiaiai.maramasaeva.com/en/dossier/not-just-sandbox AIAIAIen 2026-09-29T00:00:00.000Z An OpenAI security employee writes that a sandbox alone is not enough https://aiaiai.maramasaeva.com/dossier/astra-cancelled AIAIAInl 2026-09-29T00:00:00.000Z OpenAI schrapt de release van GPT-6.1 Astra in oktober https://aiaiai.maramasaeva.com/en/dossier/astra-cancelled AIAIAIen 2026-09-29T00:00:00.000Z OpenAI cancels the October release of GPT-6.1 Astra https://aiaiai.maramasaeva.com/dossier/pacing-frontier AIAIAInl 2026-09-29T00:00:00.000Z Meer dan 1.300 labmedewerkers vragen om een manier om AI af te remmen https://aiaiai.maramasaeva.com/en/dossier/pacing-frontier AIAIAIen 2026-09-29T00:00:00.000Z More than 1,300 lab employees ask for a way to slow AI down https://aiaiai.maramasaeva.com/dossier/intelligence-explosion-paper AIAIAInl 2026-09-29T00:00:00.000Z Hinton, Bengio en de hoofdwetenschapper van OpenAI waarschuwen voor een intelligentie-explosie https://aiaiai.maramasaeva.com/en/dossier/intelligence-explosion-paper AIAIAIen 2026-09-29T00:00:00.000Z Hinton, Bengio and OpenAI's chief scientist warn of an intelligence explosion https://aiaiai.maramasaeva.com/dossier/sonnet-5-5 AIAIAInl 2026-09-29T00:00:00.000Z Anthropic lanceert Claude Sonnet 5.5, bijna even slim als zijn topmodel https://aiaiai.maramasaeva.com/en/dossier/sonnet-5-5 AIAIAIen 2026-09-29T00:00:00.000Z Anthropic launches Claude Sonnet 5.5, nearly as smart as its top model