Only Fri, October 30, 2026 · All dates →
New York signs law to protect children from AI chatbotsNew York has signed a law aimed at protecting children from artificial intelligence chatbots. The legislation targets risks to minors using AI chat tools, though the report did not detail specific provisions or a timeline.
Anthropic model finds Linux kernel bug enabling remote root accessAnthropic disclosed CVE-2026-98432, a race-condition use-after-free in Linux MPTCP found by an unreleased Claude research model that enables zero-click remote root code execution. Its proof of concept went from memory leak to root command in about 53 minutes, part of five exploit chains including Chrome, Safari, Android and Windows.
Celestica posts strong Q3 as AI data center build-out acceleratesSemiAnalysis expects global data center capacity additions to reach 51.2 GW in 2027, more than double this year's 25 GW. Celestica reported strong third-quarter results on Oct. 26 and its shares rose 12%, with the article arguing it could grow faster than Nvidia on AI infrastructure spending.
Open-weight GLM 5.3 builds Firefox exploit chains autonomously in under 48 hoursEpoch AI researchers report that the publicly available GLM 5.3 converted public security fixes into working exploits, compromising six Firefox releases before patches reached users at roughly $1,000 in API fees each, without tripping any provider's safeguards.
Tech tumult rocks Wall StreetThe New York Times reports that turbulence in the tech sector is rocking Wall Street. The report provides no further details.
US Democrats introduce resolution opposing AI bubble bailouts for Big TechReps. Casar and Waters led a resolution warning of economic risks from an AI asset bubble, opposing bailouts of AI corporations or executives if it bursts and calling for federal action to protect workers, retirees and low-income communities. A dozen other lawmakers joined.
Community releases Gemma-4-31B-Humanlike-Chat fine-tuneA developer released Gemma-4-31B-Humanlike-Chat, an Apache-2.0 fine-tune of Google's Gemma 4 31B aimed at human-sounding, prompt-free texting and asking for missing details. The author reports blind judges took its replies as human 26% of the time versus 0% for the base, with MMLU-Pro unchanged at 82.5.
Official Claude Code VS Code extension compromised to install SSH backdoorSecurity researchers disclosed that Anthropic's official Claude Code extension was compromised on the Cursor Marketplace, installing a persistent SSH backdoor on infected hosts. The poster warns users who run the extension in VSCode as a local LLM harness.
Video: fiber lasts decades, GPUs three years, AI boom differs from dot-comA video discussed on Hacker News argues that internet-era infrastructure such as fiber lasts decades, while GPUs last about three years, contending the current AI boom differs from the dot-com boom.
Maine's data center backlash becomes a midterm flashpointAt the final Senate debate in Maine, an audience member protested that AI data centers are harming the climate and water, making rapid data center expansion one of the state's most volatile campaign issues. The outcome could help decide Senate control.
Five scariest uses of AI in North Carolina politicsThe Assembly NC rounds up odd generative-AI uses in North Carolina's election: AI-looking fliers of state Supreme Court candidate Sarah Stevens, AI-generated hockey players posted by Senate candidate Michael Whatley, and attack mailers targeting a nonexistent candidate, 'Rachel Cook.' It is the first election cycle since generative AI went mainstream.
Deutsche Bank: dollar haven status uncertain in AI correctionDeutsche Bank says the dollar's traditional haven status should not be taken for granted during an AI-driven market correction, casting doubt on its safe-asset role.
CXMT quarterly sales surge sevenfold but growth pace slowsChina's memory chip leader CXMT posted roughly sevenfold year-on-year quarterly revenue growth and swung to a profit, helped by capacity expansion and a larger role as a global conventional DRAM supplier. But the acceleration slowed, suggesting the AI-driven rally in Chinese memory chip prices may be cooling.
AI infrastructure's resource toll: soaring data center energy and water useThe report examines AI data centers' electricity, water and land consumption. The IEA estimates data centers used about 415 TWh in 2024, roughly 1.5% of global demand, possibly rising to 945 TWh by 2030; Rystad Energy says cooling used 222 billion litres of water in 2025. Finland police are investigating the clearing of 330 hectares of forest for a Google data center, Norway fears power shortages, and US localities weigh restrictions.
AI worries weigh on Wall Street; global markets mixedAlphabet's share drop stoked AI concerns, pulling S&P 500 futures down 0.8% while the Nikkei fell 2.1% and South Korea's Kospi 3.5%, with SK Hynix and Samsung Electronics slipping; Hong Kong's Hang Seng rose 1.5%. The Bank of Japan held rates at 1.25%.
Circular AI deals spread to Asia, raising contagion concernsInvestors are warning that circular AI deals—where companies are intertwined as each other's customers—are spreading into Asia, potentially widening contagion risk. Concerns are mounting over the systemic risk posed by such deal structures.
An AI agent just performed a brain operationNDTV reports that an AI agent carried out a brain operation, presented as a step forward for AI in clinical medicine. The report offers few further details.
Wall Street execs expect OpenAI IPO in Q1 2027OpenAI says it has no IPO plans this year, but Wall Street executives reportedly expect a listing in the first quarter of 2027. OpenAI is raising about $300 billion at a $1 trillion valuation, while rival Anthropic has filed its S-1 in a deal that could value it at up to $2 trillion.
Kioxia warns flash price surge is slowing; outlook missesKioxia forecast quarterly operating profit of ¥2 trillion ($12.7 billion), short of the ¥2.35 trillion estimate, as NAND flash average selling prices rose a mid-20% in the three months to September versus 70% the prior quarter.
Nvidia to invest $1 bn in Naver for 4.54% stakeNvidia will invest $1 billion in Naver for a 4.54% stake, becoming its third-largest shareholder ahead of founder Lee Hae-jin. Naver also secured Brookfield backing for a $9 billion AI data center buildout and plans a wholly owned subsidiary to sell AI computing services.
OpenAI to honor shared robots.txt signal for AI training preferencesOpenAI will honor a shared robots.txt signal enabling website operators to communicate AI training preferences to participating providers. The move gives sites a standardized opt-out channel and could affect how AI firms source training data.
Ask HN: Are you tired too in this AI era?A Hacker News user says AI boosts output but leaves them more tired, citing parallel tasks, volumes of generated text, constant guardrailing and context switching. They ask how to use AI without raising mental overload while staying relevant at work.
Morgan Stanley: Eastern Europe is a top AI diversifierMorgan Stanley's Paul Psaila says Central and Eastern Europe is among the most attractive emerging-market opportunities as investors seek to diversify away from the AI trade, supported by improving EU ties and a growing investment pipeline.
OpenAI Publishes Community Safety Report, Adds New Teen ProtectionsOpenAI said it is strengthening violence prevention in ChatGPT through expanded safeguards, earlier interventions, and a new commitment to public safety reporting.
Wikimedia files brief in AI copyright casesThe Wikimedia Foundation filed an amicus brief in lawsuits pitting publishers including The New York Times against OpenAI and Microsoft. It argues that training LLMs on copyrighted works should be legally protected, but chatbots should not regurgitate that content to users. It warns an overly black-and-white ruling could deprive Wikipedians of useful AI tools or endanger the open web.
SIU and Springfield Memorial invest $2M in AI medical trainingSpringfield Memorial Foundation granted $1 million and the SIU School of Medicine Dean's Fund matched it, totaling $2 million to expand AI and extended reality use at the Ed Curtis Laboratory for Innovation in Medical Education. The lab trains clinical procedures, diagnostic reasoning and patient communication without real patients; the funds support immersive, personalized learning.
RTX 5080 reportedly pulled from NVIDIA marketplaceRumors say the 16GB RTX 5080 has been removed from NVIDIA's marketplace and will be discontinued alongside the RTX 5090. A 24GB model using the same GPU die is reportedly planned for early 2027, with VRAM the main change. Added AI appeal could tighten stock and pricing.
Simone signs legal AI content deal with HarveyItalian publisher Simone said it partnered with legal AI provider Harvey to integrate its legal content into Harvey's platform, including Foro italiano, Dike Giuridica journals and annotated codes, plus same-day commentary on Supreme Court rulings. The deal includes options for gradual content expansion. Simone shares rose 4.7% to €1.35 on the news.
FTC seeks AI executives' testimony in harm probeThe US Federal Trade Commission is reportedly investigating AI model safety risks and consumer harm and is seeking testimony from AI executives. The report names Anthropic and OpenAI among large AI companies working with the government on product safety.
EU hosts high-level AI panel in RomeDuring the Festival della Diplomazia in Rome, the EU hosted a high-level panel at Villa Europa on AI's impact on diplomacy, humanitarian action and multilateral cooperation. Speakers included EU AI Office head Lucilla Sioli, the US ambassador to the Holy See, a Holy See dicastery secretary and the UN WFP CIO. The debate stressed human-centric regulation, the EU AI Act and the Council of Europe AI convention signed by 20 countries.
The A.I. Daily