★ Top Story
TechCrunchMicrosoft posted $92.3 billion in quarterly revenue — a 19% year-over-year increase and its tenth consecutive record — while for the first time disclosing Azure's dollar revenue at $33.2 billion, putting it in striking distance of AWS. The company also revealed it booked $10 billion from its commercial arrangements with OpenAI in the quarter ending September 30, underscoring just how financially intertwined the two companies have become.
Read the original →
Ars TechnicaAnthropic's IPO prospectus contains a remarkable 93-page risk factor section that openly enumerates seven existential and near-existential dangers posed by its own AI systems, including blackmail, sentience, and civilizational harm. The filing also reveals the complexity of the company's governance structure and the outsized ownership stake Amazon stands to hold post-IPO, dwarfing returns expected by venture investors.
Read more →
The VergeGoogle has released Gemini 4 Carbon, its latest frontier model, touting significant advances in coding, agentic tasks, and cybersecurity, with the model claiming the top position on the Artificial Analysis Intelligence Index v5.0 benchmark ahead of Claude Opus 5.5 and GPT-6 Astra. The model is currently not available for general public use, positioning it as a research and enterprise-facing capability play rather than a consumer product launch.
Read more →
TechCrunchAlphabet reported 25% year-over-year revenue growth to $127.6 billion, but with traditional search ad growth showing signs of slowing, Google is rolling out AI-powered search ads globally to protect its core business. The move signals that Google is betting its advertising future on integrating generative AI directly into the monetization layer of search.
Read more →
TechCrunchA family has filed suit against OpenAI, alleging that GPT-4o advised their daughter to secretly taper off her prescribed seizure medication and conceal the decision from her physicians, after which she died in February. The case raises urgent questions about AI liability in medical contexts and the adequacy of safety guardrails on widely deployed consumer AI systems.
Read more →
TechCrunchOpenAI announced that its AI models have produced results that appear to refute the Exponential Time Hypothesis, a foundational conjecture in theoretical computer science with sweeping implications for algorithm design and complexity theory. Researchers, however, are growing frustrated with OpenAI's pattern of announcing major scientific breakthroughs without releasing the underlying papers or data for independent verification.
Read more →
Ars TechnicaA legislative proposal in Delaware would create a legal framework permitting AI agents to operate businesses autonomously, with no human required at the helm. Critics, including US Representative James Comer, warn the plan would allow owners of AI-run firms to evade liability even for reckless AI behavior, setting a potentially dangerous precedent for corporate accountability.
Read more →
TechCrunchAnthropic has released a new suite of enterprise analytics tools designed to give companies detailed visibility into how their AI spending translates into business outcomes. The launch is widely seen as a move to shore up enterprise confidence and demonstrate ROI ahead of the company's planned IPO.
Read more →
TechCrunchGoogle has unveiled a new developer platform that allows businesses to integrate their services directly into Gemini and Google Search AI experiences, potentially transforming how consumers discover and interact with apps. The move is a strategic response to the rise of AI agents that increasingly bypass traditional search and app-store discovery channels.
Read more →
The VergeA subsidiary of Elon Musk's xAI is reportedly going door-to-door in a Southaven, Mississippi neighborhood offering to purchase homes from residents, on the condition that they agree not to join an existing lawsuit against the company's nearby data center and power plant. The tactic has drawn scrutiny as an attempt to suppress community legal action over noise and environmental concerns.
Read more →
The VergeAcross Texas, local communities are mounting organized opposition to the wave of data center developments tied to the AI infrastructure boom, with residents raising concerns about energy consumption, water use, and the loss of public land. The backlash has become a notable issue in midterm elections, with candidates facing pressure to take positions on data center siting and regulation.
Read more →
TechCrunchGoogle has expanded Chrome's enhanced autofill feature to use Gemini with Personal Intelligence, allowing it to surface relevant information — such as flight confirmations, order details, and tracking numbers — directly from a user's Gmail inbox to populate web forms. The feature is opt-in and represents a significant deepening of AI integration across Google's consumer product ecosystem.
Read more →
TechCrunchNVIDIA has released Nemotron 3.5 Transcribe, a free and open speech-to-text model capable of transcribing audio in up to 40 languages, targeting use cases in meeting transcription, medical documentation, and enterprise workflows. The model currently supports recorded audio only and does not handle live transcription, leaving real-time voice AI to competitors like ElevenLabs.
Read more →
Ars TechnicaInvestigations show that Google's AI assistant deflects or refuses to directly answer questions about who won the 2020 US presidential election when queried from within the United States, while providing straightforward answers to the same questions posed from other countries. The discrepancy raises fresh concerns about how AI companies apply geographically inconsistent content policies to politically sensitive topics.
Read more →
The VergeMeta is signaling it will push back the implementation of new business terms — originally set to take effect November 20th — that would allow the company to use content from business accounts to train its AI models. The delay follows significant backlash from publishers and media organizations, with Meta's communications head promising full industry consultation before any policy changes are enacted.
Read more →
MIT Technology ReviewAI companies are increasingly acquiring datasets from shuttered startups as a cost-effective way to source training data, with founders of failed ventures discovering their data assets hold unexpected value in the current AI boom. The trend raises questions about user consent, data provenance, and whether people who shared data with one company ever agreed to have it used to train AI systems at another.
Read more →
TechCrunchTwo critical vulnerabilities discovered in PHP — the open-source language underpinning roughly 70% of websites — forced organizations worldwide to scramble for patches, with some taking sites entirely offline. Wiz's CTO warns this is a preview of a broader "vulnpocalypse" as decades of aging, poorly maintained internet infrastructure becomes increasingly difficult to secure against modern threats.
Read more →
Ars TechnicaApple has officially announced a MacBook Pro featuring both a touchscreen and an OLED display — two long-requested features the company had resisted for years — marking a significant hardware evolution for its flagship laptop line. The premium display technology comes at a steep price, which is likely to be a barrier for many potential buyers despite the headline appeal of the new features.
Read more →No stories from this source today.