Top AI & Tech News Stories — September 30, 2026
The top AI and tech news story of September 30, 2026 was “Google’s Gemini 4 Argon safety report says it can’t rule out critical cyber and bio risk”, covered by 21 outlets (23 articles), according to Skillenai's news index. 20 stories ranked across 169 articles.
Last updated · 20 stories · 169 articles
Get the top AI and tech news in your inbox each evening. No account needed, unsubscribe anytime.
Google’s Gemini 4 Argon safety report says it can’t rule out critical cyber and bio risk
21 outlets · 23 articles · 137 total since September 20, 2026 · Google, Gemini 4 Argon, OpenAI, Anthropic, Google DeepMind
Announced 13 Oct 2026 · Sources checked What did Google publish? When Google made Gemini 4 Argon generally available on 13 October, its model card twice said a full Safety & Security Report was “coming soon.” That evening Google DeepMind posted “Developing Gemini 4 Argon…
Read at thenextweb.com →First reported by geeky-gadgets.comMore coverage: constellationr.com · nytimes.com · the-decoder.com · the-decoder.de · trendingtopics.eu
Sol Nearly Eclipses Astra: OpenAI’s GPT-6.1 Sol is DevDay’s star, joined by Dots and new ChatGPT integrations
38 outlets · 40 articles · 145 total since September 28, 2026 · OpenAI, dots, ChatGPT, GPT-6 Astra, Meta
OpenAI’s latest mid-tier model lands within a point of its flagship at less than a quarter of the cost per task on Artificial Analysis’ Intelligence Index. OpenAI upgraded Sol one week after that model's debut, but its Astra model’s upgrade was withdrawn days before its planned…
Read at me.pcmag.com →First reported by digestai.newsMore coverage: how2shout.com · geeky-gadgets.com · searchenginejournal.com · ciol.com · webpronews.com
Exploitbench mostra che un salto di una generazione può trasformare un crash in un exploit completo
9 outlets · 10 articles · 15 total since September 30, 2026 · anthropic, exploit development, GLM-5.3, Claude Mythos Preview, ExploitBench
Un modello frontier è stato messo davanti a un bug noto nel motore V8 di Chrome, insieme alla relativa patch, e gli è stato chiesto di costruire l’exploit. Nel caso di Claude Mythos Preview, il risultato è stato molto più vicino a quello di un ricercatore umano di sicurezza che…
Read at the-decoder.de →First reported by trendingtopics.euMore coverage: scmp.com · trendingtopics.eu · punto-informatico.it · aiinsiders.net · tomshardware.com
Is Mark Zuckerberg the new Trump-Whisperer?
11 outlets · 13 articles · 39 total since September 29, 2026 · anthropic, Meta, OpenAI, NVIDIA, Google
At the White House state dinner for Chinese President Xi Jinping on Sept. 24, Mark Zuckerberg was seated beside House Speaker Mike Johnson. The Meta chief executive talked to the Louisiana Republican about regulatory concerns around AI, people familiar with the conversation…
Read at wire.fourthweb.ai →First reported by techechelon.comMore coverage: onlinemarketing.de · techxmedia.com · news.symplexia.com · news.google.com · computersweden.se
GPT-6.1 Sol su GitHub Copilot: il nuovo modello agentico è classificato “Critical” per la cybersecurity
15 outlets · 16 articles · 88 total since July 9, 2026 · OpenAI, GPT-6 Sol, GPT-6 Astra, GPT-6.1 Sol, codex
GitHub ha reso disponibile in GitHub Copilot GPT-6.1 Sol, il nuovo modello di OpenAI pensato esplicitamente per agentic coding e workflow da terminale. La notizia, in sé, sarebbe una delle tante aggiunte al model picker di Copilot che si susseguono ogni poche settimane. Ma c’è…
Read at trendingtopics.eu →First reported by cnbc.comMore coverage: drweb.de · ai4business.it · techbriefly.com · innovatopia.jp · ai.watch.impress.co.jp
McDonald’s AI Pricing Push: Recommendations or Quiet Control Over the Big Mac?
3 outlets · 3 articles · 5 total since September 30, 2026 · McDonald's, Big Mac, Wendy's, dynamic pricing, algorithmic pricing
Two McDonald’s locations sit just two miles apart in Fresno, California. One sells a Big Mac for $5.69. The other charges $6.89. That 21% gap surfaced last month on the company’s mobile app. It captures a larger tension now gripping the fast-food giant. McDonald’s has spent…
More coverage: trendingtopics.eu · webpronews.com
OpenAI and Synopsys partner to build "GPT-Synopsys" for autonomous chip design — specialized AI model will operate EDA tools, allowing engineers to deliver more sophisticated chips faster - Tom's Hardware
11 outlets · 12 articles · 29 total since September 28, 2026 · synopsys, OpenAI, electronic design automation (EDA), GPT-Synopsys, chip design
OpenAI and Synopsys partner to build "GPT-Synopsys" for autonomous chip design — specialized AI model will operate EDA tools, allowing engineers to deliver more sophisticated chips faster Tom's Hardware
Read at interestingengineering.com →First reported by futurumresearch.comMore coverage: the-decoder.de · nerds.xyz · convergedigest.com · news.google.com · news.synopsys.com
Daily TEA – What Does Meta's Muse Know About Your Friends?
4 outlets · 4 articles · 60 total since September 12, 2026 · Meta, MUSE, Amazon, Apple, OpenAI
Muse's private dossiers, AI agents in finance, chatbot election advice, inference economics, and the UK's safety data fight Hello, dear TEA-mates! Here is what you need to know today. 1. 🔍 Meta's Muse Maps Users' Relationships and Personal Habits Meta's Muse personal assistant…
Read at eu.techcrunch.com →First reported by cybernewscentre.comMore coverage: sqmagazine.co.uk · tomshardware.com · itsecuritynews.info
SpaceXAI Weighs Four-Tier Pricing For Grok And X Users
1 outlet · 1 article · 3 total since September 30, 2026 · SpaceXAI, bloomberg, Grok, X, subscription pricing
Read the full analysis: SpaceXAI Weighs Four-Tier Pricing For Grok And X Users on ThorstenMeyerAI.com TL;DR Bloomberg reports that xAI is considering a four-tier pricing model for users of Grok and X. The report does not establish that the plans are approved, and no prices…
Musk's Dot.com Redirects To Grok As OpenAI Launches Dots
7 outlets · 7 articles · 27 total since September 29, 2026 · OpenAI, dots, ChatGPT, GPT-6.1 Sol, GPT-6 Astra
SpaceXAI, the Elon Musk company that traded as xAI until SpaceX acquired it and renamed it in July, owns the domain dot.com and has it redirecting to the download page for Grok Bot, its own AI agent product. The redirect surfaced on 29 September, the day OpenAI announced Dots, a…
Read at me.pcmag.com →First reported by itdoeswhatnow.comMore coverage: technobezz.com · nexchron.com · techbriefly.com · computerhoy.20minutos.es · blogspan.net
We tested America.gov, Trump’s new AI ‘concierge.’ Here’s what it can do.
3 outlets · 3 articles · 5 total since September 29, 2026 · America.gov, America.gov, Office of Management and Budget, GSA, login.gov
The chatbot promises to help users with passports, Medicare and more. It can provide information, but it can’t take actions for you yet.
Read at meta.slashdot.org →First reported by govciomedia.comMore coverage: aiassemblage.org · theworldofai.org
Pentagon personnel breach — undetected for months — renews cyber standard scrutiny
2 outlets · 2 articles · 12 total since September 30, 2026 · Credit monitoring, CNN, United States, Defense Manpower Data Center, Defense Manpower Data Center (DMDC) - Consortium
A Pentagon breach that exposed sensitive personnel information for months before discovery is raising questions yet again about how federal agencies are able to successfully detect unauthorized access to the records they hold on millions of individuals. Introduction to Malware…
Read at meta.slashdot.org →First reported by cisovoice.comMore coverage: cisovoice.com
OpenAI Holds Back GPT-6.1 Astra After Safety Tests Miss Bar
4 outlets · 4 articles · 19 total since September 28, 2026 · OpenAI, GPT-6 Astra, UK AI Security Institute, GPT-6.1 Astra, GPT-5.6 Sol
OpenAI Holds Back GPT-6.1 Astra After Safety Tests Miss Bar - OpenAI delayed GPT-6.1 Astra after testing found it did not meet the company’s safety bar for staying within scope and authorization. [1] - OpenAI safety chief Saachi Jain said the model was more persistent at…
Read at sealion-lavender-ar2n.squarespace.com →First reported by itdoeswhatnow.comMore coverage: punto-informatico.it · ct.nl · theworldofai.org
High-severity OpenSSL flaw can leak memory or crash programs
5 outlets · 5 articles · openssl, wolfSSL, SSL/TLS, DTLS, patch management
The vulnerability arises in the Datagram Transport Layer Security (DTLS) protocol, a variant of TLS used for UDP traffic.
Read at scmagazine.com →First reported by techinsider.comMore coverage: techinsider.com · news4hackers.com · techmaniacs.com · news.mlab.sh
President Tokayev: The "Washington Model" for AI Could Form the Basis of an International Treaty Under UN Auspices
5 outlets · 5 articles · 23 total since September 29, 2026 · Meta, OpenAI, anthropic, Donald Trump, NVIDIA
Kazakhstan's President Kassym-Jomart Tokayev has proposed using the voluntary artificial intelligence accord signed by President Donald Trump and major U.S. technology companies as a model for international adoption. He calls this approach the "Washington Model" and says it…
Read at androidheadlines.com →First reported by itdoeswhatnow.comMore coverage: impli.me · metacurity.com · cnbc.com · theworldofai.org
OpenAI Is Hitting the Brakes on Wall Street Debut and Safety Is the Reason
11 outlets · 12 articles · 87 total since September 28, 2026 · OpenAI, Saachi Jain, wall street journal, GPT-6.1 Astra, ChatGPT
Wall Street will have to wait a bit longer for the biggest public offering in tech. Speaking to reporters after his DevDay keynote, OpenAI CEO Sam Altman made it clear that the company will not go public until it can make solid, confident safety guarantees about its artificial…
Read at me.pcmag.com →First reported by itdoeswhatnow.comMore coverage: cybersecitalia.it · basic-tutorials.com · swisscybersecurity.net · jarvis-ai.cz · autonainews.com
Ghee haw
2 outlets · 2 articles · Windows 95, Android, Apple ecosystem, iPad, iPhone 14 Pro Max
The White House praises data centers while announcing a self-policing accord for artificial intelligence companies. / Semafor President Trump says government employees should refer to AI as “super intelligence,” because “the word super is the best word of all.” / The Verge [$]…
More coverage: themorningnews.org
How Capital One evaluates, governs its AI agents
1 outlet · 1 article · AI agents, LLM, Hallucination detection, Human-in-the-loop, evaluation metrics
How Capital One evaluates, governs its AI agents Larry Dignan Wed, 30 Sep 2026 - 13:29 Larry Dignan Editor in Chief of Constellation Insights Constellation Research Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial…
Notorious Spectre CPU vulnerability returns hitting JIT engines via side channel attacks
4 outlets · 4 articles · 5 total since September 29, 2026 · Spectre-v2, speculative execution, Branch Target Reuse (BTR), Vrije Universiteit Amsterdam (VU), Scuola Superiore Sant’Anna
Notorious Spectre CPU vulnerability returns hitting JIT engines via side channel attacks Branch Target Reuse is the latest application of the dreaded Spectre flaw - New Spectre variant exploits processor prediction behavior in just-in-time compilers - Researchers demonstrated…
Read at apcmag.com →First reported by news.mlab.shMore coverage: ictsecuritymagazine.com · theregister.co.uk · news.mlab.sh
CoreWeave Fully Connected 2026: Forge Brings Frontier Lab RL to Every AI Team
2 outlets · 2 articles · 6 total since September 30, 2026 · coreweave, CoreWeave Forge, observability, NVIDIA, San Francisco
The post CoreWeave Fully Connected 2026: Forge Brings Frontier Lab RL to Every AI Team appeared first on Futurum . Analyst(s): Brendan Burke, Nick Patience Publication Date: October 6, 2026 At Fully Connected 2026, CoreWeave launched Forge as a unified environment for the AI…
More coverage: aicybr.com
Get the top AI and tech news in your inbox
One email each evening with the day's biggest stories. No account needed, unsubscribe anytime.
How this was computed
Articles in the Skillenai news index are grouped into event clusters at enrichment time (hybrid embedding/BM25/entity candidate search plus an LLM join-vs-new judge). Stories are ranked by the authority of the articles published on 2026-09-30 (UTC) — the sum of their outlet and author authority — so a multi-day story only ranks on the strength of that day's coverage. Article and outlet counts are likewise scoped to the day. The highlighted link is the day's highest-authority article in each cluster.