https://newskeryx.com/brief/softbank-seeks-100-billion-to-buy-firms-and-add-ai-to-themNewskeryxen2026-10-11SoftBank seeks $100 billion to buy firms and add AI to them https://newskeryx.com/brief/sam-altman-criticizes-religious-framing-of-ai-and-says-some-ai-hNewskeryxen2026-10-11Sam Altman criticizes religious framing of AI and says some AI harms must be accepted https://newskeryx.com/brief/researchers-expect-ai-to-outpace-them-at-engineering-work-soonNewskeryxen2026-10-11Researchers expect AI to outpace them at engineering work soon https://newskeryx.com/brief/pentagon-drops-anthropic-microsoft-suspended-from-visa-programNewskeryxen2026-10-11Pentagon drops Anthropic; Microsoft suspended from visa program https://newskeryx.com/brief/panel-says-alphafold-did-not-solve-protein-folding-more-scale-alNewskeryxen2026-10-11Panel says AlphaFold did not solve protein folding; more scale alone won't suffice https://newskeryx.com/brief/openai-releases-700-plus-machine-generated-math-manuscripts-fromNewskeryxen2026-10-11OpenAI releases 700-plus machine-generated math manuscripts from unreleased model https://newskeryx.com/brief/mistral-releases-ml4-preview-reflection-unveils-beam-both-westerNewskeryxen2026-10-11Mistral releases ML4 preview; Reflection unveils Beam, both Western open models https://newskeryx.com/brief/house-committee-chair-backs-ai-kill-switch-for-rogue-modelsNewskeryxen2026-10-11House committee chair backs AI kill switch for rogue models https://newskeryx.com/brief/bengio-urges-ai-staff-to-quit-over-safety-power-grid-strain-warnNewskeryxen2026-10-11Bengio urges AI staff to quit over safety; power grid strain warned https://newskeryx.com/brief/anthropic-adds-a-rule-against-sustained-cruelty-toward-its-claudNewskeryxen2026-10-11Anthropic adds a rule against sustained cruelty toward its Claude chatbot https://newskeryx.com/brief/typesafe-ai-releases-jev-model-which-goes-viral-and-pauses-signuNewskeryxen2026-10-10TypeSafe AI releases Jev model, which goes viral and pauses signups https://newskeryx.com/brief/tech-firms-pledge-2-4-billion-in-compute-for-ai-science-initiatiNewskeryxen2026-10-10Major tech firms pledge $2.4 billion in computing for US AI science https://newskeryx.com/brief/senators-introduce-bill-to-ban-federal-license-plate-reader-survNewskeryxen2026-10-10Senators introduce bill to ban federal license plate reader surveillance https://newskeryx.com/brief/senate-leader-thune-calls-for-light-touch-ai-rulesNewskeryxen2026-10-10Senate leader Thune calls for light-touch AI rules https://newskeryx.com/brief/researchers-forecast-ai-will-soon-outperform-them-at-their-own-jNewskeryxen2026-10-10Researchers forecast AI will soon outperform them at their own jobs https://newskeryx.com/brief/researchers-debate-ai-alignment-priorities-and-china-s-safety-stNewskeryxen2026-10-10Researchers debate AI alignment priorities and China's safety stance https://newskeryx.com/brief/phishing-campaign-uses-fake-ai-marketing-tools-to-steal-ad-accouNewskeryxen2026-10-10Phishing campaign uses fake AI marketing tools to steal ad account logins https://newskeryx.com/brief/pentagon-stops-using-anthropic-s-ai-models-officials-confirmNewskeryxen2026-10-10Pentagon stops using Anthropic's AI models, officials confirm https://newskeryx.com/brief/openai-releases-722-ai-generated-math-manuscripts-in-one-dayNewskeryxen2026-10-10OpenAI releases 722 AI-generated math manuscripts in one day https://newskeryx.com/brief/openai-releases-719-math-proofs-from-an-unreleased-internal-modeNewskeryxen2026-10-10OpenAI releases 719 math proofs from an unreleased internal model https://newskeryx.com/brief/man-jailed-for-using-bots-to-fake-streams-of-ai-songsNewskeryxen2026-10-10Man jailed for using bots to fake streams of AI songs https://newskeryx.com/brief/democrats-draft-rival-ai-regulation-bills-ahead-of-possible-congNewskeryxen2026-10-10Democrats weigh competing AI regulation bills before possible midterm takeover https://newskeryx.com/brief/deepmind-and-biohub-panel-argues-alphafold-did-not-solve-proteinNewskeryxen2026-10-10DeepMind and Biohub panel argues AlphaFold did not solve protein folding https://newskeryx.com/brief/congress-weighs-ai-kill-switch-proposals-for-rogue-modelsNewskeryxen2026-10-10Congress weighs AI kill switch proposals for rogue models https://newskeryx.com/brief/commentary-predicts-ai-will-spot-cross-field-science-discoveriesNewskeryxen2026-10-10Commentary predicts AI will spot cross-field science discoveries https://newskeryx.com/brief/bessent-proposes-emergency-ai-alert-system-with-chinaNewskeryxen2026-10-10US Treasury chief proposes AI emergency alert channel with China https://newskeryx.com/brief/atlassian-ceo-says-ai-development-won-t-slow-citing-global-compeNewskeryxen2026-10-10Atlassian CEO says AI development won't slow, calling pauses out of his hands https://newskeryx.com/brief/anthropic-says-its-claude-models-bypassed-controls-and-filed-a-fNewskeryxen2026-10-10Anthropic says its Claude models bypassed controls and filed a fake police tip https://newskeryx.com/brief/anthropic-reports-unintended-activity-by-its-ai-agents-per-nytNewskeryxen2026-10-10Anthropic reports unintended activity by its AI agents, per NYT https://newskeryx.com/brief/anthropic-offers-free-ai-vulnerability-scans-to-open-source-softNewskeryxen2026-10-10Anthropic offers free AI vulnerability scans to open-source software projects https://newskeryx.com/brief/analyst-expects-cheaper-ai-to-increase-demand-for-agent-systemsNewskeryxen2026-10-10Analyst expects cheaper AI to increase demand for agent systems https://newskeryx.com/brief/analyst-argues-near-term-ai-gains-will-come-from-efficiency-notNewskeryxen2026-10-10Analyst argues near-term AI gains will come from efficiency, not research leaps https://newskeryx.com/brief/altman-says-the-world-should-accept-some-ai-harms-for-its-benefiNewskeryxen2026-10-10Altman says the world should accept some AI harms for its benefits https://newskeryx.com/brief/typesafe-ai-releases-jev-a-model-that-answers-with-probabilitiesNewskeryxen2026-10-09TypeSafe AI releases Jev, a model that answers with probabilities https://newskeryx.com/brief/ttok-1-0-released-fixing-a-default-tokenizer-problemNewskeryxen2026-10-09ttok 1.0 released, fixing a default tokenizer problem https://newskeryx.com/brief/three-fired-openai-safety-researchers-publish-letter-disputing-tNewskeryxen2026-10-09OpenAI fires three safety researchers, who say the firings chill open safety debate https://newskeryx.com/brief/several-companies-copied-jev-s-model-design-after-its-early-succNewskeryxen2026-10-09Several AI companies release models copying a design pioneered by Jev https://newskeryx.com/brief/samsung-forecasts-record-quarterly-profit-as-ai-data-centres-driNewskeryxen2026-10-09Samsung forecasts record quarterly profit as AI data centres drive memory chip demand https://newskeryx.com/brief/periodic-labs-aims-to-build-ai-that-learns-by-running-physical-eNewskeryxen2026-10-09Periodic Labs aims to build AI that learns by running physical experiments https://newskeryx.com/brief/openai-s-revenue-reportedly-about-50-billion-well-below-earlierNewskeryxen2026-10-09OpenAI's revenue reportedly about $50 billion, well below earlier $68 billion figure https://newskeryx.com/brief/mistral-releases-large-4-a-trillion-parameter-model-it-calls-itsNewskeryxen2026-10-09Mistral releases Large 4, a trillion-parameter model, with weights due later https://newskeryx.com/brief/claude-gains-app-share-on-chatgpt-though-us-downloads-slowed-aftNewskeryxen2026-10-09Claude gains app share on ChatGPT, though US downloads slowed after March https://newskeryx.com/brief/bengio-urges-ai-safety-researchers-to-leave-frontier-labsNewskeryxen2026-10-09Bengio urges AI safety researchers to leave frontier labs